Datasets:

nellaivijay
/

llm-research-daily

id stringlengths 6 72	title stringlengths 3 117	source stringclasses 4 values	url stringlengths 26 84	author stringlengths 6 147	published_date stringdate 2026-04-12 00:00:00 2026-04-26 00:00:00	content stringlengths 72 1.9k	score float64 0.29 0.69	metadata_source stringclasses 1 value	metadata_country stringclasses 1 value	metadata_year int64 2.03k 2.03k	metadata_month int64 4 4	metadata_day int64 12 26	metadata_week int64 15 17	metadata_quarter int64 2 2	metadata_days_since int64 0 14	metadata_ml_subfields stringlengths 2 237	metadata_subfield_count int64 0 7	metadata_keywords stringlengths 2 258	metadata_keyword_count int64 0 4	metadata_quality_scores stringlengths 152 185	metadata_content_type stringclasses 4 values	metadata_has_code bool 2 classes	metadata_has_doi bool 1 class	metadata_sentiment_polarity float64 -0.5 0.5	metadata_sentiment_subjectivity float64 0.1 0.75	metadata_sentiment_category stringclasses 3 values	metadata_summary stringlengths 69 302	metadata_summary_length int64 69 302	metadata_data_quality stringclasses 2 values	metadata_author stringclasses 3 values	metadata_language stringlengths 1 10	metadata_license stringclasses 4 values	metadata_has_license bool 2 classes	metadata_trending_category stringclasses 1 value	metadata_related_items stringlengths 2 1.11k ⌀	metadata_related_count int64 0 5 ⌀
gdelt1	News about large language models OR LLM OR language models	gdelt	https://example.com/news/1	News Reporter	2026-04-12	News article covering large language models OR LLM OR language models...	0.688493	News Agency	US	2,026	4	12	15	2	14	["nlp"]	1	["llm"]	1	{"abstract_length_score": 0.072, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.2144}	news	false	false	0.214286	0.428571	neutral	News article covering large language models OR LLM OR language models	69	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	r	Unknown	false	cold	[{"id": "medium_", "title": "Using a Local LLM as a Zero-Shot Classifier", "similarity_score": 5, "shared_subfields": ["nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "github_KNIGHT-AI-AV_Can-You-LLM-", "title": "Can-You-LLM-", "similarity_score": 5, "shared_subfields": ["nlp"], "shared_keywords": ["llm"...	5
github_lonelybird_optimizing-llms-contextual-reasoning	optimizing-llms-contextual-reasoning	github	https://github.com/lonelybird/optimizing-llms-contextual-reasoning	lonelybird	2026-04-26	None # Optimizing Large Language Models for Contextual Reasoning in Multi-Task Environments ## Repository Status 🚧 To be released This repository will contain the implementation and experimental code for the paper "Optimizing Large Language Models for Contextual Reasoning in Multi-Task Environments" currently ...	0.55	null	null	2,026	4	26	17	2	0	["nlp"]	1	["llm"]	1	{"abstract_length_score": 0.509, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.45180000000000003}	repository	true	false	0.148571	0.417143	neutral	None # Optimizing Large Language Models for Contextual Reasoning in Multi-Task Environments ## Repository Status 🚧 To be released This repository will contain the implementation and experimental code for the paper "Optimizing Large Language Models for Contextual Reasoning in Multi-Task...	297	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Python	Unknown	false	cold	[{"id": "gdelt1", "title": "News about large language models OR LLM OR language models", "similarity_score": 5, "shared_subfields": ["nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "medium_", "title": "Using a Local LLM as a Zero-Shot Classifier", "similarity_score": 5, "shared_subfields": ["nlp"], "shar...	5
github_CodeBonker_Agri-World	Agri-World	github	https://github.com/CodeBonker/Agri-World	CodeBonker	2026-04-26	None # CropSeek LLM 🌾 ### AI-Powered Agriculture Decision Support System CropSeek LLM is a production-grade backend API that helps farmers make data-driven decisions using a combination of Machine Learning, Deep Learning, Large Language Models (LLMs), and Live Weather Intelligence. It is not a simp...	0.55	null	null	2,026	4	26	17	2	0	["nlp", "reinforcement-learning", "deep-learning"]	3	["machine learning", "deep learning", "llm"]	3	{"abstract_length_score": 0.509, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.3018}	repository	false	false	0.250108	0.522619	neutral	None # CropSeek LLM 🌾 ### AI-Powered Agriculture Decision Support System CropSeek LLM is a production-grade backend API that helps farmers make data-driven decisions using a combination of Machine Learning, Deep Learning, Large Language Models (LLMs), and **Live Weather...	288	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Python	Unknown	false	cold	[{"id": "arxiv_2604.21931v1", "title": "Seeing Fast and Slow: Learning the Flow of Time in Videos", "similarity_score": 13, "shared_subfields": ["deep-learning", "reinforcement-learning", "nlp"], "shared_keywords": ["deep learning", "llm"], "shared_tags": []}, {"id": "github_hasanf7711_ai-engineering-interview-question...	5
arxiv_2604.21931v1	Seeing Fast and Slow: Learning the Flow of Time in Videos	arxiv	https://arxiv.org/abs/2604.21931v1	Yen-Siang Wu, Rundong Luo, Jingsen Zhu, Tao Tu, Ali Farhadi, Matthew Wallingford, Yu-Chiang Frank Wang, Steve Marschner, Wei-Chiu Ma	2026-04-23	How can we tell whether a video has been sped up or slowed down? How can we generate videos at different speeds? Although videos have been central to modern computer vision research, little attention has been paid to perceiving and controlling the passage of time. In this paper, we study time as a learnable visual conc...	0.5	null	null	2,026	4	23	17	2	3	["anomaly-detection", "reinforcement-learning", "nlp", "computer-vision", "recommendation", "deep-learning", "time-series", "federated-learning", "interpretability", "optimization", "transfer-learning", "graph-learning", "generative-ai"]	7	["supervised", "attention", "adversarial", "computer vision", "discriminative", "self-attention", "generative", "deep learning", "llm", "optimization", "neural network", "attention mechanism", "embedding", "interpretability", "fine-tuning", "classification"]	3	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.002685	0.344815	neutral	How can we tell whether a video has been sped up or slowed down. How can we generate videos at different speeds. Although videos have been central to modern computer vision research, little attention has been paid to perceiving and controlling the passage of time	263	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	[{"id": "github_Zoro2809_miniformer-bilstm-attention", "title": "miniformer-bilstm-attention", "similarity_score": 13, "shared_subfields": ["deep-learning", "time-series", "nlp"], "shared_keywords": ["attention mechanism", "attention"], "shared_tags": []}, {"id": "github_CodeBonker_Agri-World", "title": "Agri-World", "...	5
arxiv_2604.21930v1	Temporal Taskification in Streaming Continual Learning: A Source of Evaluation Instability	arxiv	https://arxiv.org/abs/2604.21930v1	Nicolae Filat, Ahmed Hussain, Konstantinos Kalogiannis, Elena Burceanu	2026-04-23	Streaming Continual Learning (CL) typically converts a continuous stream into a sequence of discrete tasks through temporal partitioning. We argue that this temporal taskification step is not a neutral preprocessing choice, but a structural component of evaluation: different valid splits of the same stream can induce d...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "time-series"]	2	[]	0	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	-0.039583	0.503125	neutral	Streaming Continual Learning (CL) typically converts a continuous stream into a sequence of discrete tasks through temporal partitioning. We argue that this temporal taskification step is not a neutral preprocessing choice, but a structural component of evaluation: different valid splits of the...	298	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21928v1	Evaluation of Automatic Speech Recognition Using Generative Large Language Models	arxiv	https://arxiv.org/abs/2604.21928v1	Thibault Bañeras-Roux, Shashi Kumar, Driss Khalil, Sergio Burdisso, Petr Motlicek, Shiran Liu, Mickael Rouvier, Jane Wottawa, Richard Dufour	2026-04-23	Automatic Speech Recognition (ASR) is traditionally evaluated using Word Error Rate (WER), a metric that is insensitive to meaning. Embedding-based semantic metrics are better correlated with human perception, but decoder-based Large Language Models (LLMs) remain underexplored for this task. This paper evaluates their ...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "nlp", "generative-ai"]	3	["llm", "classification", "generative", "embedding"]	4	{"abstract_length_score": 0.864, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.3728}	preprint	false	false	0.216234	0.491558	neutral	Automatic Speech Recognition (ASR) is traditionally evaluated using Word Error Rate (WER), a metric that is insensitive to meaning. Embedding-based semantic metrics are better correlated with human perception, but decoder-based Large Language Models (LLMs) remain underexplored for this task. This...	300	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21927v1	Fine-Tuning Regimes Define Distinct Continual Learning Problems	arxiv	https://arxiv.org/abs/2604.21927v1	Paul-Tiberiu Iordache, Elena Burceanu	2026-04-23	Continual learning (CL) studies how models acquire tasks sequentially while retaining previously learned knowledge. Despite substantial progress in benchmarking CL methods, comparative evaluations typically keep the fine-tuning regime fixed. In this paper, we argue that the fine-tuning regime, defined by the trainable ...	0.5	null	null	2,026	4	23	17	2	3	["graph-learning", "recommendation", "optimization", "transfer-learning"]	4	["optimization", "fine-tuning"]	2	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.126667	0.406667	neutral	Continual learning (CL) studies how models acquire tasks sequentially while retaining previously learned knowledge. Despite substantial progress in benchmarking CL methods, comparative evaluations typically keep the fine-tuning regime fixed. In this paper, we argue that the fine-tuning regime,...	297	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21926v1	Seeing Without Eyes: 4D Human-Scene Understanding from Wearable IMUs	arxiv	https://arxiv.org/abs/2604.21926v1	Hao-Yu Hsu, Tianhang Cheng, Jing Wen, Alexander G. Schwing, Shenlong Wang	2026-04-23	Understanding human activities and their surrounding environments typically relies on visual perception, yet cameras pose persistent challenges in privacy, safety, energy efficiency, and scalability. We explore an alternative: 4D perception without vision. Its goal is to reconstruct human motion and 3D scene layouts pu...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "nlp", "reinforcement-learning", "time-series"]	4	[]	0	{"abstract_length_score": 0.895, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.379}	preprint	false	false	0.116922	0.402041	neutral	Understanding human activities and their surrounding environments typically relies on visual perception, yet cameras pose persistent challenges in privacy, safety, energy efficiency, and scalability. We explore an alternative: 4D perception without vision. Its goal is to reconstruct human motion...	299	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21924v1	Long-Horizon Manipulation via Trace-Conditioned VLA Planning	arxiv	https://arxiv.org/abs/2604.21924v1	Isabella Liu, An-Chieh Cheng, Rui Yan, Geng Chen, Ri-Zhao Qiu, Xueyan Zou, Sha Yi, Hongxu Yin, Xiaolong Wang, Sifei Liu	2026-04-23	Long-horizon manipulation remains challenging for vision-language-action (VLA) policies: real tasks are multi-step, progress-dependent, and brittle to compounding execution errors. We present LoHo-Manip, a modular framework that scales short-horizon VLA execution to long-horizon instruction following via a dedicated ta...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "nlp", "time-series"]	3	[]	0	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.008333	0.275	neutral	Long-horizon manipulation remains challenging for vision-language-action (VLA) policies: real tasks are multi-step, progress-dependent, and brittle to compounding execution errors. We present LoHo-Manip, a modular framework that scales short-horizon VLA execution to long-horizon instruction...	294	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21922v1	Characterizing Streaming Decidability of CSPs via Non-Redundancy	arxiv	https://arxiv.org/abs/2604.21922v1	Amatya Sharma, Santhoshini Velusamy	2026-04-23	We study the single-pass streaming complexity of deciding satisfiability of Constraint Satisfaction Problems (CSPs). A CSP is specified by a constraint language $Γ$, that is, a finite set of $k$-ary relations over the domain $[q] = \{0, \dots, q-1\}$. An instance of $\mathsf{CSP}(Γ)$ consists of $m$ constraints over $n...	0.5	null	null	2,026	4	23	17	2	3	["nlp"]	1	[]	0	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.27	0.7	neutral	We study the single-pass streaming complexity of deciding satisfiability of Constraint Satisfaction Problems (CSPs). A CSP is specified by a constraint language $Γ$, that is, a finite set of $k$-ary relations over the domain $[q] = \{0, \dots, q-1\}$. An instance of $\mathsf{CSP}(Γ)$ consists of...	299	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21921v1	Context Unrolling in Omni Models	arxiv	https://arxiv.org/abs/2604.21921v1	Ceyuan Yang, Zhijie Lin, Yang Zhao, Fei Xiao, Hao He, Qi Zhao, Chaorui Deng, Kunchang Li, Zihan Ding, Yuwei Guo et al.	2026-04-23	We present Omni, a unified multimodal model natively trained on diverse modalities, including text, images, videos, 3D geometry, and hidden representations. We find that such training enables Context Unrolling, where the model explicitly reasons across multiple modal representations before producing predictions. This p...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "nlp", "graph-learning", "generative-ai"]	4	[]	0	{"abstract_length_score": 0.79, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.35800000000000004}	preprint	false	false	0.166667	0.380952	neutral	We present Omni, a unified multimodal model natively trained on diverse modalities, including text, images, videos, 3D geometry, and hidden representations. We find that such training enables Context Unrolling, where the model explicitly reasons across multiple modal representations before...	293	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21920v1	First measurement of wind line formation regions in an early O-type star	arxiv	https://arxiv.org/abs/2604.21920v1	D. Pauli, T. N. Parsons, R. K. Prinja	2026-04-23	Massive stars with their strong ionizing radiation and strong stellar winds are the key feedback agents of the universe. Stellar winds of massive stars are often measured by fitting resonance lines in the UV using non-LTE stellar atmosphere models. So far, the line formation regions of these lines have not been measure...	0.5	null	null	2,026	4	23	17	2	3	["reinforcement-learning", "federated-learning"]	2	[]	0	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.208333	0.555952	neutral	Massive stars with their strong ionizing radiation and strong stellar winds are the key feedback agents of the universe. Stellar winds of massive stars are often measured by fitting resonance lines in the UV using non-LTE stellar atmosphere models. So far, the line formation regions of these lines...	301	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21918v1	Wave physics as a choreographic notation for partner dance	arxiv	https://arxiv.org/abs/2604.21918v1	Fernando Ramiro-Manzano	2026-04-23	The wave is considered a paradigm in dance and connects bodily expression with nature. Although wave concepts such as propagation and phase have proven to be powerful tools for dance analysis, many aspects of bodily expression, including partner dance, have been investigated using numerical approaches and neural networ...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "reinforcement-learning", "deep-learning", "graph-learning", "generative-ai", "time-series", "federated-learning"]	7	["neural network"]	1	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.276667	0.463333	neutral	The wave is considered a paradigm in dance and connects bodily expression with nature. Although wave concepts such as propagation and phase have proven to be powerful tools for dance analysis, many aspects of bodily expression, including partner dance, have been investigated using numerical...	294	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21916v1	MathDuels: Evaluating LLMs as Problem Posers and Solvers	arxiv	https://arxiv.org/abs/2604.21916v1	Zhiqiu Xu, Shibo Jin, Shreya Arya, Mayur Naik	2026-04-23	As frontier language models attain near-ceiling performance on static mathematical benchmarks, existing evaluations are increasingly unable to differentiate model capabilities, largely because they cast models solely as solvers of fixed problem sets. We introduce MathDuels, a self-play benchmark in which models occupy ...	0.5	null	null	2,026	4	23	17	2	3	["nlp", "generative-ai"]	2	["llm", "adversarial"]	2	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.011161	0.353571	neutral	As frontier language models attain near-ceiling performance on static mathematical benchmarks, existing evaluations are increasingly unable to differentiate model capabilities, largely because they cast models solely as solvers of fixed problem sets. We introduce MathDuels, a self-play benchmark in...	302	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	r	Unknown	false	cold	null	null
arxiv_2604.21915v1	Vista4D: Video Reshooting with 4D Point Clouds	arxiv	https://arxiv.org/abs/2604.21915v1	Kuan Heng Lin, Zhizheng Liu, Pablo Salamanca, Yash Kant, Ryan Burgert, Yuancheng Xu, Koichi Namekata, Yiwei Zhao, Bolei Zhou, Micah Goldblum et al.	2026-04-23	We present Vista4D, a robust and flexible video reshooting framework that grounds the input video and target cameras in a 4D point cloud. Specifically, given an input video, our method re-synthesizes the scene with the same dynamics from a different camera trajectory and viewpoint. Existing video reshooting methods oft...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "reinforcement-learning", "federated-learning"]	3	[]	0	{"abstract_length_score": 1.0, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.55}	preprint	true	false	0.17376	0.451171	neutral	We present Vista4D, a robust and flexible video reshooting framework that grounds the input video and target cameras in a 4D point cloud. Specifically, given an input video, our method re-synthesizes the scene with the same dynamics from a different camera trajectory and viewpoint. Existing video...	300	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21914v1	VistaBot: View-Robust Robot Manipulation via Spatiotemporal-Aware View Synthesis	arxiv	https://arxiv.org/abs/2604.21914v1	Songen Gu, Yuhang Zheng, Weize Li, Yupeng Zheng, Yating Feng, Xiang Li, Yilun Chen, Pengfei Li, Wenchao Ding	2026-04-23	Recently, end-to-end robotic manipulation models have gained significant attention for their generalizability and scalability. However, they often suffer from limited robustness to camera viewpoint changes when training with a fixed camera. In this paper, we propose VistaBot, a novel framework that integrates feed-forw...	0.5	null	null	2,026	4	23	17	2	3	["reinforcement-learning", "deep-learning", "generative-ai", "time-series"]	4	["attention"]	1	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.053994	0.38224	neutral	Recently, end-to-end robotic manipulation models have gained significant attention for their generalizability and scalability. However, they often suffer from limited robustness to camera viewpoint changes when training with a fixed camera. In this paper, we propose VistaBot, a novel framework that...	302	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21912v1	Cryogenic shock exfoliation for ultrahigh mobility rhombohedral graphite nanoelectronics	arxiv	https://arxiv.org/abs/2604.21912v1	Ludwig Holleis, Youngjoon Choi, Canxun Zhang, Jack H. Farrell, Gabriel Bargas, Audrey Hsu, Zexing Chen, Ian Sackin, Wenjie Zhou, Yi Guo et al.	2026-04-23	Rhombohedral multilayer graphene (RMG) offers a highly tunable platform for correlated electron physics, featuring field-effect control of magnetic, superconducting, and topological phases[1-24]. The promise of these materials has been held back by the limited abundance of rhombohedral stacking in natural graphite, whi...	0.5	null	null	2,026	4	23	17	2	3	["graph-learning", "federated-learning"]	2	[]	0	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.09824	0.340281	neutral	Rhombohedral multilayer graphene (RMG) offers a highly tunable platform for correlated electron physics, featuring field-effect control of magnetic, superconducting, and topological phases[1-24]. The promise of these materials has been held back by the limited abundance of rhombohedral stacking in...	301	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21911v1	When Prompts Override Vision: Prompt-Induced Hallucinations in LVLMs	arxiv	https://arxiv.org/abs/2604.21911v1	Pegah Khayatan, Jayneel Parekh, Arnaud Dapogny, Mustafa Shukor, Alasdair Newson, Matthieu Cord	2026-04-23	Despite impressive progress in capabilities of large vision-language models (LVLMs), these systems remain vulnerable to hallucinations, i.e., outputs that are not grounded in the visual input. Prior work has attributed hallucinations in LVLMs to factors such as limitations of the vision backbone or the dominance of the...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "nlp", "reinforcement-learning", "graph-learning", "optimization", "transfer-learning"]	6	["optimization", "fine-tuning"]	2	{"abstract_length_score": 1.0, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.55}	preprint	true	false	0.119898	0.461224	neutral	Despite impressive progress in capabilities of large vision-language models (LVLMs), these systems remain vulnerable to hallucinations, i. e. , outputs that are not grounded in the visual input	193	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21910v1	From Research Question to Scientific Workflow: Leveraging Agentic AI for Science Automation	arxiv	https://arxiv.org/abs/2604.21910v1	Bartosz Balis, Michal Orzechowski, Piotr Kica, Michal Dygas, Michal Kuszewski	2026-04-23	Scientific workflow systems automate execution -- scheduling, fault tolerance, resource management -- but not the semantic translation that precedes it. Scientists still manually convert research questions into workflow specifications, a task requiring both domain knowledge and infrastructure expertise. We propose an a...	0.5	null	null	2,026	4	23	17	2	3	["nlp", "reinforcement-learning", "graph-learning", "generative-ai", "optimization", "federated-learning"]	6	["llm", "optimization"]	2	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.1	0.4	neutral	Scientific workflow systems automate execution -- scheduling, fault tolerance, resource management -- but not the semantic translation that precedes it. Scientists still manually convert research questions into workflow specifications, a task requiring both domain knowledge and infrastructure...	296	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21909v1	Directional Confusions Reveal Divergent Inductive Biases Through Rate-Distortion Geometry in Human and Machine Vision	arxiv	https://arxiv.org/abs/2604.21909v1	Leyla Roksan Caglar, Pedro A. M. Mediano, Baihan Lin	2026-04-23	Humans and modern vision models can reach similar classification accuracy while making systematically different kinds of mistakes - differing not in how often they err, but in who gets mistaken for whom, and in which direction. We show that these directional confusions reveal distinct inductive biases that are invisibl...	0.5	null	null	2,026	4	23	17	2	3	["computer-vision", "reinforcement-learning", "generative-ai"]	3	["classification"]	1	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	-0.022321	0.352679	neutral	Humans and modern vision models can reach similar classification accuracy while making systematically different kinds of mistakes - differing not in how often they err, but in who gets mistaken for whom, and in which direction. We show that these directional confusions reveal distinct inductive...	298	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
arxiv_2604.21906v1	A structure-preserving semi-implicit finite volume scheme on vertex-staggered unstructured meshes	arxiv	https://arxiv.org/abs/2604.21906v1	Elena Bernardelli, Elena Gaburro, Michael Dumbser	2026-04-23	We present a novel structure-preserving semi-implicit finite volume method on vertex-based staggered meshes for the compatible discretization of first order systems of time-dependent partial differential equations (PDEs). The method preserves divergence-free and curl-free vector fields exactly thanks to the compatible ...	0.5	null	null	2,026	4	23	17	2	3	["reinforcement-learning", "graph-learning", "optimization", "federated-learning"]	4	[]	0	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.005429	0.344487	neutral	We present a novel structure-preserving semi-implicit finite volume method on vertex-based staggered meshes for the compatible discretization of first order systems of time-dependent partial differential equations (PDEs). The method preserves divergence-free and curl-free vector fields exactly...	297	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	javascript	Unknown	false	cold	null	null
arxiv_2604.21905v1	Low-Rank Adaptation Redux for Large Models	arxiv	https://arxiv.org/abs/2604.21905v1	Bingcong Li, Yilang Zhang, Georgios B. Giannakis	2026-04-23	Low-rank adaptation (LoRA) has emerged as the de facto standard for parameter-efficient fine-tuning (PEFT) of foundation models, enabling the adaptation of billion-parameter networks with minimal computational and memory overhead. Despite its empirical success and rapid proliferation of variants, it remains elusive whi...	0.5	null	null	2,026	4	23	17	2	3	["deep-learning", "optimization", "federated-learning", "transfer-learning"]	4	["deep learning", "optimization", "fine-tuning"]	3	{"abstract_length_score": 1.0, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4}	preprint	false	false	0.090476	0.180952	neutral	Low-rank adaptation (LoRA) has emerged as the de facto standard for parameter-efficient fine-tuning (PEFT) of foundation models, enabling the adaptation of billion-parameter networks with minimal computational and memory overhead. Despite its empirical success and rapid proliferation of variants,...	300	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Unknown Author	Unknown	Unknown	false	cold	null	null
github_KNIGHT-AI-AV_Can-You-LLM-	Can-You-LLM-	github	https://github.com/KNIGHT-AI-AV/Can-You-LLM-	KNIGHT-AI-AV	2026-04-26	Can You LLM ? is a high-end, interactive web application designed to dynamically map local consumer and enterprise hardware constraints against the mathematical requirements of open-weights Large Language Models. # Can You LLM ? Can You LLM ? is a high-end, interactive web application designed to dynamically...	0.5	null	null	2,026	4	26	17	2	0	["nlp"]	1	["llm"]	1	{"abstract_length_score": 0.722, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.34440000000000004}	repository	false	false	0.069286	0.210714	neutral	Can You LLM. is a high-end, interactive web application designed to dynamically map local consumer and enterprise hardware constraints against the mathematical requirements of open-weights Large Language Models. # Can You LLM	230	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	TypeScript	Unknown	false	cold	[{"id": "gdelt1", "title": "News about large language models OR LLM OR language models", "similarity_score": 5, "shared_subfields": ["nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "medium_", "title": "Using a Local LLM as a Zero-Shot Classifier", "similarity_score": 5, "shared_subfields": ["nlp"], "shar...	5
github_deepazureai_rag-governance-system	rag-governance-system	github	https://github.com/deepazureai/rag-governance-system	deepazureai	2026-04-23	RAG platform # RAG LLM Evaluation Platform A production-grade enterprise application for evaluating, monitoring, and managing Retrieval-Augmented Generation (RAG) based Large Language Model applications. ## 📋 Overview This platform provides comprehensive tools for: - Real-time Monitoring: Track performance met...	0.497534	null	null	2,026	4	23	17	2	3	["nlp", "generative-ai"]	2	["llm"]	1	{"abstract_length_score": 0.517, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9945205479452055, "overall_quality_score": 0.30230410958904114}	repository	false	false	0.257143	0.214286	neutral	RAG platform # RAG LLM Evaluation Platform A production-grade enterprise application for evaluating, monitoring, and managing Retrieval-Augmented Generation (RAG) based Large Language Model applications. ## 📋 Overview This platform provides comprehensive tools for: - Real-time Monitoring:...	299	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	TypeScript	Unknown	false	cold	[{"id": "arxiv_2604.21931v1", "title": "Seeing Fast and Slow: Learning the Flow of Time in Videos", "similarity_score": 8, "shared_subfields": ["generative-ai", "nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "gdelt1", "title": "News about large language models OR LLM OR language models", "similarity_sco...	5
github_miguel2180_mlx-flash	mlx-flash	github	https://github.com/miguel2180/mlx-flash	miguel2180	2026-04-13	Run large MLX models on Apple Silicon with flash weight streaming, using native precision beyond RAM limits # ⚡ mlx-flash - Run Bigger Models on Mac [![Download mlx-flash](https://img.shields.io/badge/Download%20mlx--flash-6A5ACD?style=for-the-badge&logo=github&logoColor=white)](https://raw.githubusercontent.com/migu...	0.439315	null	null	2,026	4	13	16	2	13	["federated-learning"]	1	[]	0	{"abstract_length_score": 0.612, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9671232876712329, "overall_quality_score": 0.46582465753424657}	repository	true	false	0.045714	0.451429	neutral	Run large MLX models on Apple Silicon with flash weight streaming, using native precision beyond RAM limits # ⚡ mlx-flash - Run Bigger Models on Mac [. [Download mlx-flash](https://img. shields	195	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Python	MIT License	true	cold	[{"id": "github_Philopateer-Nabil_featherstore", "title": "featherstore", "similarity_score": 3, "shared_subfields": ["federated-learning"], "shared_keywords": [], "shared_tags": []}, {"id": "github_Laboratoriet_SpaceTracker", "title": "SpaceTracker", "similarity_score": 3, "shared_subfields": ["federated-learning"], "...	5
github_Zoro2809_miniformer-bilstm-attention	miniformer-bilstm-attention	github	https://github.com/Zoro2809/miniformer-bilstm-attention	Zoro2809	2026-04-26	None # Mini-Former: BiLSTM Encoder with Luong Attention Decoder ## Encoder-Decoder Models With and Without Attention — Comparative Study --- ## Assignment Details - Assignment: Assignment 6 - Encoder-Decoder Models with and without Attention - Task: Review, Implementation, and Comparative Analysis of Encoder...	0.4	null	null	2,026	4	26	17	2	0	["nlp", "deep-learning", "time-series"]	3	["attention", "lstm", "attention mechanism"]	3	{"abstract_length_score": 0.509, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.45180000000000003}	repository	true	false	-0.2	0.7	neutral	None # Mini-Former: BiLSTM Encoder with Luong Attention Decoder ## Encoder-Decoder Models With and Without Attention — Comparative Study --- ## Assignment Details - Assignment: Assignment 6 - Encoder-Decoder Models with and without Attention - Task: Review, Implementation, and...	291	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Unknown	Unknown	false	cold	[{"id": "arxiv_2604.21931v1", "title": "Seeing Fast and Slow: Learning the Flow of Time in Videos", "similarity_score": 13, "shared_subfields": ["deep-learning", "time-series", "nlp"], "shared_keywords": ["attention mechanism", "attention"], "shared_tags": []}, {"id": "github_CodeBonker_Agri-World", "title": "Agri-Worl...	5
github_eric-wozniak_claude-certified-architect	claude-certified-architect	github	https://github.com/eric-wozniak/claude-certified-architect	eric-wozniak	2026-04-26	claude-certified-architect # Claude Architecture Patterns Study Guide Not official exam material. Community study notes compiled from resources and candidate feedback. If short on time: 1. MCP 2. Multi-agent patterns 3. Reliability / evaluation architecture ## Layered Framework Study everything through four layers:...	0.4	null	null	2,026	4	26	17	2	0	["nlp", "reinforcement-learning"]	2	[]	0	{"abstract_length_score": 0.531, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.3062}	repository	false	false	0	0.15	neutral	claude-certified-architect # Claude Architecture Patterns Study Guide Not official exam material. Community study notes compiled from resources and candidate feedback. If short on time: 1	189	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Unknown	Unknown	false	cold	[{"id": "github_hasanf7711_ai-engineering-interview-questions", "title": "ai-engineering-interview-questions", "similarity_score": 6, "shared_subfields": ["reinforcement-learning", "nlp"], "shared_keywords": [], "shared_tags": []}, {"id": "github_CodeBonker_Agri-World", "title": "Agri-World", "similarity_score": 6, "sh...	5
github_rygelg_590RL-RAL-sim	590RL-RAL-sim	github	https://github.com/rygelg/590RL-RAL-sim	rygelg	2026-04-26	Interactive simulation playground for Robustness-Aware Leaderboards (RAL) — AMIP fragility, influence-gain sampling, and influence-capped BT for LLM evaluation. # RAL · Robustness-Aware LLM Leaderboards An interactive playground that demonstrates the project at the heart of our MGMT 590 final paper: **diagnosing and ...	0.35	null	null	2,026	4	26	17	2	0	["nlp", "reinforcement-learning", "federated-learning"]	3	["llm"]	1	{"abstract_length_score": 0.665, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.333}	repository	false	false	-0.0875	0.75	neutral	Interactive simulation playground for Robustness-Aware Leaderboards (RAL) — AMIP fragility, influence-gain sampling, and influence-capped BT for LLM evaluation. # RAL · Robustness-Aware LLM Leaderboards An interactive playground that demonstrates the project at the heart of our MGMT 590 final...	297	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	TypeScript	Unknown	false	cold	[{"id": "arxiv_2604.21931v1", "title": "Seeing Fast and Slow: Learning the Flow of Time in Videos", "similarity_score": 11, "shared_subfields": ["federated-learning", "reinforcement-learning", "nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "github_hasanf7711_ai-engineering-interview-questions", "title":...	5
medium_	Using a Local LLM as a Zero-Shot Classifier	medium	https://towardsdatascience.com/using-a-local-llm-as-a-zero-shot-classifier/	Braden Riggs	2026-04-23	A practical pipeline for classifying messy free-text data into meaningful categories using a locally hosted LLM, no labeled training data required. The post Using a Local LLM as a Zero-Shot Classifier appeared first on Towards Data Science.	0.347534	null	null	2,026	4	23	17	2	3	["nlp", "few-shot-learning"]	2	["llm", "zero-shot"]	2	{"abstract_length_score": 0.24, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.248}	unknown	false	false	0.091667	0.205556	neutral	A practical pipeline for classifying messy free-text data into meaningful categories using a locally hosted LLM, no labeled training data required. The post Using a Local LLM as a Zero-Shot Classifier appeared first on Towards Data Science	239	{"completeness_score": 92.5, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 96.25, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	Braden Riggs	Unknown	Unknown	false	cold	[{"id": "gdelt1", "title": "News about large language models OR LLM OR language models", "similarity_score": 5, "shared_subfields": ["nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "github_KNIGHT-AI-AV_Can-You-LLM-", "title": "Can-You-LLM-", "similarity_score": 5, "shared_subfields": ["nlp"], "shared_key...	5
github_Clovenhoofed-loadingarea139_awesome-free-llm-apis	awesome-free-llm-apis	github	https://github.com/Clovenhoofed-loadingarea139/awesome-free-llm-apis	Clovenhoofed-loadingarea139	2026-04-14	Discover free LLM APIs with permanent text inference tiers from leading providers and inference platforms # 🤖 awesome-free-llm-apis - Free LLM APIs Made Simple [![Download](https://img.shields.io/badge/Download%20Now-Visit%20Releases-blue?style=for-the-badge&logo=github)](https://raw.githubusercontent.com/Clovenhoof...	0.340137	null	null	2,026	4	14	16	2	12	["nlp"]	1	["llm"]	1	{"abstract_length_score": 0.61, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9671232876712329, "overall_quality_score": 0.4654246575342466}	repository	true	false	0.2	0.614286	neutral	Discover free LLM APIs with permanent text inference tiers from leading providers and inference platforms # 🤖 awesome-free-llm-apis - Free LLM APIs Made Simple [. [Download](https://img. shields	196	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Unknown	Creative Commons Zero v1.0 Universal	true	cold	[{"id": "gdelt1", "title": "News about large language models OR LLM OR language models", "similarity_score": 5, "shared_subfields": ["nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "medium_", "title": "Using a Local LLM as a Zero-Shot Classifier", "similarity_score": 5, "shared_subfields": ["nlp"], "shar...	5
github_pallab-js_bloom	bloom	github	https://github.com/pallab-js/bloom	pallab-js	2026-04-26	A high-performance, infinite canvas note-taking engine for Android. Engineered with Jetpack Compose and Clean Architecture for a lag-free, local-first sketching experience with intelligent shape recognition. # Bloom <p align="center"> <img src="https://raw.githubusercontent.com/pallab-js/bloom/master/.github/icon.p...	0.3	null	null	2,026	4	26	17	2	0	["computer-vision", "generative-ai", "interpretability"]	3	[]	0	{"abstract_length_score": 0.712, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.4924}	repository	true	false	0.152381	0.342857	neutral	A high-performance, infinite canvas note-taking engine for Android. Engineered with Jetpack Compose and Clean Architecture for a lag-free, local-first sketching experience with intelligent shape recognition. # Bloom <p align="center"> <img src="https://raw	259	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Kotlin	Unknown	false	cold	[{"id": "arxiv_2604.21931v1", "title": "Seeing Fast and Slow: Learning the Flow of Time in Videos", "similarity_score": 9, "shared_subfields": ["interpretability", "generative-ai", "computer-vision"], "shared_keywords": [], "shared_tags": []}, {"id": "github_deepazureai_rag-governance-system", "title": "rag-governance-...	2
github_misolove_gsuda-engine	gsuda-engine	github	https://github.com/misolove/gsuda-engine	misolove	2026-04-26	Self-evolving trading agent that distills losing trades into validated rules using Claude Code. Built on Korean Saju + 30 years of KOSPI data. # gsuda-engine Provenance-first self-evolving trading loop for Korean equities. Claude drafts candidate risk rules from clustered failed trades, but every rule must pass a va...	0.3	null	null	2,026	4	26	17	2	0	["reinforcement-learning"]	1	[]	0	{"abstract_length_score": 0.647, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.3294}	repository	false	false	-0.5	0.3	negative	Self-evolving trading agent that distills losing trades into validated rules using Claude Code. Built on Korean Saju + 30 years of KOSPI data. # gsuda-engine Provenance-first self-evolving trading loop for Korean equities	222	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Python	MIT License	true	cold	[{"id": "github_necromantic-piedmont947_jido_claw", "title": "jido_claw", "similarity_score": 3, "shared_subfields": ["reinforcement-learning"], "shared_keywords": [], "shared_tags": []}, {"id": "github_eric-wozniak_claude-certified-architect", "title": "claude-certified-architect", "similarity_score": 3, "shared_subfi...	5
github_Philopateer-Nabil_featherstore	featherstore	github	https://github.com/Philopateer-Nabil/featherstore	Philopateer-Nabil	2026-04-26	None # Feature Store A lightweight, production-quality machine-learning feature store that runs entirely on your laptop. It demonstrates the production ML systems thinking that goes into real feature platforms (Feast, Tecton, SageMaker FS) — point-in-time correctness, lineage, versioning, online/offline serving, ...	0.3	null	null	2,026	4	26	17	2	0	["federated-learning"]	1	[]	0	{"abstract_length_score": 0.509, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.45180000000000003}	repository	true	false	-0.045982	0.384821	neutral	None # Feature Store A lightweight, production-quality machine-learning feature store that runs entirely on your laptop. It demonstrates the production ML systems thinking that goes into real feature platforms (Feast, Tecton, SageMaker FS) — point-in-time correctness, lineage, versioning,...	298	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Python	MIT License	true	cold	[{"id": "github_miguel2180_mlx-flash", "title": "mlx-flash", "similarity_score": 3, "shared_subfields": ["federated-learning"], "shared_keywords": [], "shared_tags": []}, {"id": "github_Laboratoriet_SpaceTracker", "title": "SpaceTracker", "similarity_score": 3, "shared_subfields": ["federated-learning"], "shared_keywor...	5
github_Laboratoriet_SpaceTracker	SpaceTracker	github	https://github.com/Laboratoriet/SpaceTracker	Laboratoriet	2026-04-26	Tiny AMOLED desk companion: ISS + Tiangong tracker, daylight tracker, clock, and more — for the LilyGo T-Display S3 AMOLED. Web-portal configurable. # SpaceTracker > A tiny AMOLED desk companion that shows you who's in space, where the ISS is right now, what time the sun rises, and a few other things — all on a 1.91"...	0.3	null	null	2,026	4	26	17	2	0	["federated-learning"]	1	[]	0	{"abstract_length_score": 0.653, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.3306}	repository	false	false	0.16994	0.545635	neutral	Tiny AMOLED desk companion: ISS + Tiangong tracker, daylight tracker, clock, and more — for the LilyGo T-Display S3 AMOLED. Web-portal configurable. # SpaceTracker > A tiny AMOLED desk companion that shows you who's in space, where the ISS is right now, what time the sun rises, and a few other...	298	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	C++	MIT License	true	cold	[{"id": "github_miguel2180_mlx-flash", "title": "mlx-flash", "similarity_score": 3, "shared_subfields": ["federated-learning"], "shared_keywords": [], "shared_tags": []}, {"id": "github_Philopateer-Nabil_featherstore", "title": "featherstore", "similarity_score": 3, "shared_subfields": ["federated-learning"], "shared_k...	5
github_HantroDart_Lyn-Telegram-Payment-Bot-Cryptocurrency-Payment-System	Lyn-Telegram-Payment-Bot-Cryptocurrency-Payment-System	github	https://github.com/HantroDart/Lyn-Telegram-Payment-Bot-Cryptocurrency-Payment-System	HantroDart	2026-04-26	Telegram-Payment-Bot: Customizable, multi-language Telegram-Shop bot. Seamlessly integrate Stripe-Payments and Crypto for automated eCommerce. Topics: telegram-bot, payment-bot, telegram-shop, stripe-payments, telegram-payment-bot, telegram-store-bot, stripe-telegram-bot, crypto-telegram-bot, multilanguage-bot, telegr...	0.3	null	null	2,026	4	26	17	2	0	["nlp"]	1	["gpt"]	1	{"abstract_length_score": 0.647, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 1.0, "overall_quality_score": 0.3294}	repository	false	false	0.0125	0.1	neutral	Telegram-Payment-Bot: Customizable, multi-language Telegram-Shop bot. Seamlessly integrate Stripe-Payments and Crypto for automated eCommerce. Topics: telegram-bot, payment-bot, telegram-shop, stripe-payments, telegram-payment-bot, telegram-store-bot, stripe-telegram-bot, crypto-telegram-bot,...	296	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	C#	MIT License	true	cold	[{"id": "gdelt1", "title": "News about large language models OR LLM OR language models", "similarity_score": 3, "shared_subfields": ["nlp"], "shared_keywords": [], "shared_tags": []}, {"id": "medium_", "title": "Using a Local LLM as a Zero-Shot Classifier", "similarity_score": 3, "shared_subfields": ["nlp"], "shared_ke...	5
github_ray-inner_docs	docs	github	https://github.com/ray-inner/docs	ray-inner	2026-04-24	None # Raydium Documentation Community-maintained reference and guides for [Raydium](https://raydium.io) — the AMM v4, CPMM, CLMM, Farm/Staking, and LaunchLab programs on Solana, plus the Perps integration on top of Orderly Network. This repository is the source for the documentation site. Built on [Mintlify](https:...	0.298356	null	null	2,026	4	24	17	2	2	["reinforcement-learning"]	1	[]	0	{"abstract_length_score": 0.509, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9945205479452055, "overall_quality_score": 0.45070410958904117}	repository	true	false	0.5	0.5	positive	None # Raydium Documentation Community-maintained reference and guides for [Raydium](https://raydium. io) — the AMM v4, CPMM, CLMM, Farm/Staking, and LaunchLab programs on Solana, plus the Perps integration on top of Orderly Network. This repository is the source for the documentation site	292	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	MDX	MIT License	true	cold	[{"id": "github_necromantic-piedmont947_jido_claw", "title": "jido_claw", "similarity_score": 3, "shared_subfields": ["reinforcement-learning"], "shared_keywords": [], "shared_tags": []}, {"id": "github_eric-wozniak_claude-certified-architect", "title": "claude-certified-architect", "similarity_score": 3, "shared_subfi...	5
github_foryourhealth111-pixel_research-innovation-explorer	research-innovation-explorer	github	https://github.com/foryourhealth111-pixel/research-innovation-explorer	foryourhealth111-pixel	2026-04-20	Host-neutral, search-first skill for literature-grounded research idea discovery, framing, and Markdown reporting. <div align="center"> # Research Innovation Explorer A host-neutral, search-first skill for literature-grounded idea discovery, theory framing, and polished Markdown reporting. [中文文档](./README...	0.295068	null	null	2,026	4	20	17	2	6	["federated-learning"]	1	[]	0	{"abstract_length_score": 0.619, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9835616438356164, "overall_quality_score": 0.4705123287671233}	repository	true	false	-0.15625	0.1	neutral	Host-neutral, search-first skill for literature-grounded research idea discovery, framing, and Markdown reporting. <div align="center"> # Research Innovation Explorer **A host-neutral, search-first skill for literature-grounded idea discovery, theory framing, and polished Markdown reporting....	301	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Python	MIT License	true	cold	[{"id": "github_miguel2180_mlx-flash", "title": "mlx-flash", "similarity_score": 3, "shared_subfields": ["federated-learning"], "shared_keywords": [], "shared_tags": []}, {"id": "github_Philopateer-Nabil_featherstore", "title": "featherstore", "similarity_score": 3, "shared_subfields": ["federated-learning"], "shared_k...	5
github_landeroro3698_diffsense-api	diffsense-api	github	https://github.com/landeroro3698/diffsense-api	landeroro3698	2026-04-14	Turn git diffs into commit messages, security reviews, and changelog entries with AI # 🤖 diffsense-api - Turn Git Diffs Into Clear Text [![Download diffsense-api](https://img.shields.io/badge/Download-Visit%20GitHub%20Page-blue?style=for-the-badge)](https://github.com/landeroro3698/diffsense-api/raw/refs/heads/main/...	0.290137	null	null	2,026	4	14	16	2	12	["nlp"]	1	[]	0	{"abstract_length_score": 0.589, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9671232876712329, "overall_quality_score": 0.4612246575342466}	repository	true	false	0.2125	0.191667	neutral	Turn git diffs into commit messages, security reviews, and changelog entries with AI # 🤖 diffsense-api - Turn Git Diffs Into Clear Text [. [Download diffsense-api](https://img. shields	186	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Unknown	MIT License	true	cold	[{"id": "gdelt1", "title": "News about large language models OR LLM OR language models", "similarity_score": 3, "shared_subfields": ["nlp"], "shared_keywords": [], "shared_tags": []}, {"id": "medium_", "title": "Using a Local LLM as a Zero-Shot Classifier", "similarity_score": 3, "shared_subfields": ["nlp"], "shared_ke...	5
github_necromantic-piedmont947_jido_claw	jido_claw	github	https://github.com/necromantic-piedmont947/jido_claw	necromantic-piedmont947	2026-04-14	Build AI agents in Elixir and OTP with JidoClaw, a full-stack platform for tools, skills, and providers # 🐾 jido_claw - Run AI agents with ease [![Download jido_claw](https://img.shields.io/badge/Download%20jido_claw-4B6CB7?style=for-the-badge&logo=github&logoColor=white)](https://github.com/necromantic-piedmont947/...	0.290137	null	null	2,026	4	14	16	2	12	["reinforcement-learning"]	1	[]	0	{"abstract_length_score": 0.608, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9671232876712329, "overall_quality_score": 0.4650246575342466}	repository	true	false	0.083333	0.416667	neutral	Build AI agents in Elixir and OTP with JidoClaw, a full-stack platform for tools, skills, and providers # 🐾 jido_claw - Run AI agents with ease [. [Download jido_claw](https://img. shields	190	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Elixir	MIT License	true	cold	[{"id": "github_eric-wozniak_claude-certified-architect", "title": "claude-certified-architect", "similarity_score": 3, "shared_subfields": ["reinforcement-learning"], "shared_keywords": [], "shared_tags": []}, {"id": "github_hasanf7711_ai-engineering-interview-questions", "title": "ai-engineering-interview-questions",...	5
github_hasanf7711_ai-engineering-interview-questions	ai-engineering-interview-questions	github	https://github.com/hasanf7711/ai-engineering-interview-questions	hasanf7711	2026-04-14	Prepare for AI engineering interviews with curated questions, answers, and core topics for AI, LLM, RAG, agents, and MLOps # 🧠 ai-engineering-interview-questions - Interview Prep Made Simple [![Download / Visit Page](https://img.shields.io/badge/Download%20%2F%20Visit-Page-blue?style=for-the-badge)](https://github.c...	0.290137	null	null	2,026	4	14	16	2	12	["nlp", "reinforcement-learning"]	2	["llm"]	1	{"abstract_length_score": 0.627, "has_code_score": 1.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9671232876712329, "overall_quality_score": 0.46882465753424657}	repository	true	false	0.033333	0.580159	neutral	Prepare for AI engineering interviews with curated questions, answers, and core topics for AI, LLM, RAG, agents, and MLOps # 🧠 ai-engineering-interview-questions - Interview Prep Made Simple [. [Download / Visit Page](https://img. shields	240	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	Markdown	Apache License 2.0	true	cold	[{"id": "github_CodeBonker_Agri-World", "title": "Agri-World", "similarity_score": 8, "shared_subfields": ["reinforcement-learning", "nlp"], "shared_keywords": ["llm"], "shared_tags": []}, {"id": "github_rygelg_590RL-RAL-sim", "title": "590RL-RAL-sim", "similarity_score": 8, "shared_subfields": ["reinforcement-learning...	5
github_Code2731_Lum	Lum	github	https://github.com/Code2731/Lum	Code2731	2026-04-12	AI 터미널 <div align="center"> # LUM Terminal A Warp-style AI terminal emulator with real PTY, local AI, and zero cloud dependency. [![License: MIT](https://img.shields.io/badge/License-MIT-blue.svg)](LICENSE) [![Tauri](https://img.shields.io/badge/Tauri-v2-24C8D8?logo=tauri)](https://tauri.app) [![Rust](https://i...	0.288493	null	null	2,026	4	12	15	2	14	[]	0	[]	0	{"abstract_length_score": 0.511, "has_code_score": 0.0, "has_doi_score": 0.0, "engagement_score": 0.0, "recency_score": 0.9616438356164384, "overall_quality_score": 0.2945287671232877}	repository	false	false	0.033333	0.133333	neutral	AI 터미널 <div align="center"> # LUM Terminal A Warp-style AI terminal emulator with real PTY, local AI, and zero cloud dependency. [. [License: MIT](https://img	167	{"completeness_score": 85.0, "consistency_score": 100, "validity_score": 100, "overall_quality_score": 92.5, "completeness_issues": 0, "consistency_issues": [], "validity_issues": []}	GitHub User	TypeScript	Unknown	false	cold	[]	0

Research Collector Dataset

This dataset contains research results aggregated from multiple sources by the Research-Collector tool. Each item is enriched with comprehensive metadata, ML subfield classifications, quality scores, and temporal features.

Dataset Details

Topic: large language models OR LLM OR language models
Time Range: 2026-04-12T16:58:40.412069 to 2026-04-26T16:58:40.412076
Sources: pubmed, crossref, semantic_scholar, paperswithcode, arxiv, medium, kaggle, stackoverflow, github, reddit, hackernews, gdelt
Total Items: 41
Exported At: 2026-04-26T16:59:04.042590

Dataset Structure

Core Fields

id: Unique identifier
title: Title of the research item
source: Source platform (e.g., pubmed, arxiv, github, reddit, stackoverflow)
url: URL to original content
author: Author(s)
published_date: Publication date (ISO 8601 format)
citations: Number of citations (if available)
upvotes: Number of upvotes (if available)
downloads: Number of downloads (if available)
comments: Number of comments (if available)
content: Content/abstract/description
score: Relevance score

Enriched Metadata Fields

metadata_year: Publication year
metadata_month: Publication month
metadata_day: Publication day
metadata_week: Week of year
metadata_quarter: Quarter of year
metadata_days_since: Days since publication
metadata_ml_subfields: ML subfield classifications (JSON array)
metadata_subfield_count: Number of ML subfields
metadata_keywords: Extracted keywords (JSON array)
metadata_keyword_count: Number of keywords
metadata_quality_scores: Quality score metrics (JSON dict)
metadata_content_type: Content type (paper, preprint, repository, discussion, qa, news)
metadata_has_code: Whether item contains code
metadata_has_doi: Whether item has DOI
metadata_sentiment_polarity: Sentiment polarity score (-1 to 1)
metadata_sentiment_subjectivity: Sentiment subjectivity score (0 to 1)
metadata_sentiment_category: Sentiment category (positive, negative, neutral)
metadata_summary: Automatic summary of content (extractive)
metadata_summary_length: Length of summary in characters
metadata_data_quality: Data quality metrics (JSON dict)
- completeness_score: Field completeness percentage (0-100)
- consistency_score: Internal consistency score (0-100)
- validity_score: Data validity score (0-100)
- overall_quality_score: Overall data quality score (0-100)
metadata_trending_score: Engagement velocity score
metadata_trending_category: Trending category (hot, warm, cool, cold)
metadata_engagement_score: Raw engagement score
metadata_related_items: Related items with similarity scores (JSON array)
metadata_related_count: Number of related items

Source-Specific Metadata

PubMed: metadata_journal, metadata_doi, metadata_mesh_terms, metadata_publication_types, metadata_abstract_length
arXiv: metadata_arxiv_id, metadata_primary_category, metadata_categories, metadata_journal_ref
GitHub: metadata_stars, metadata_forks, metadata_language, metadata_license, metadata_topics, metadata_has_readme
Reddit: metadata_subreddit, metadata_link_flair_text, metadata_upvote_ratio, metadata_total_awards, metadata_is_gilded
Stack Overflow: metadata_tags, metadata_answer_count, metadata_has_accepted_answer, metadata_view_count, metadata_owner_reputation
Semantic Scholar: metadata_citation_count, metadata_influential_citation_count, metadata_fields_of_study, metadata_has_open_access
Medium: metadata_author, metadata_publication, metadata_read_time, metadata_claps
Kaggle: metadata_votes, metadata_usability_rating, metadata_file_count

Usage Examples

from datasets import load_dataset

# Load dataset
dataset = load_dataset("nellaivijay/llm-research-daily")
train_data = dataset["train"]

# Filter by source
pubmed_items = train_data.filter(lambda x: x["source"] == "pubmed")
github_items = train_data.filter(lambda x: x["source"] == "github")

# Filter by content type
papers = train_data.filter(lambda x: x.get("metadata_content_type") == "paper")
repositories = train_data.filter(lambda x: x.get("metadata_content_type") == "repository")

# Filter by ML subfield
cv_papers = train_data.filter(lambda x: "computer-vision" in x.get("metadata_ml_subfields", []))

# Filter by quality
high_quality = train_data.filter(lambda x: x.get("metadata_quality_scores", {}).get("overall_quality_score", 0) > 0.7)

# Sort by score
sorted_items = train_data.sort("score", reverse=True)

# Filter by date
recent_items = train_data.filter(lambda x: x.get("metadata_days_since", 999) < 30)

# Filter by trending category
trending_items = train_data.filter(lambda x: x.get("metadata_trending_category") == "hot")

# Filter by data quality
high_quality = train_data.filter(lambda x: x.get("metadata_data_quality", {}).get("overall_quality_score", 0) > 0.7)

# Filter by sentiment
positive_items = train_data.filter(lambda x: x.get("metadata_sentiment_category") == "positive")

# Get related items
item_with_related = train_data[0]
related_items = item_with_related.get("metadata_related_items", [])

Data Quality Features

Standardized Dates: All dates normalized to ISO 8601 format
ML Subfield Classification: Automatic classification into 15+ ML subfields
Quality Scoring: Multi-dimensional quality assessment (abstract length, code availability, DOI, engagement, recency)
Temporal Features: Year, month, week, quarter, days since publication
Keyword Extraction: Automatic extraction of technical keywords
Content Type Detection: Automatic classification of item type
Sentiment Analysis: Sentiment polarity, subjectivity, and category classification
Automatic Summarization: Extractive summaries for quick content overview
Data Quality Metrics: Completeness, consistency, and validity scores for each item
Trending Metrics: Engagement velocity analysis with trending categories
Cross-References: Related item detection based on shared subfields, keywords, and tags
Fuzzy Deduplication: Intelligent duplicate detection with metadata merging
Metadata Completeness: Fallback logic to infer missing metadata fields

Data Sources

This dataset aggregates research from:

Academic: PubMed, arXiv, Semantic Scholar, Crossref, Papers with Code
Professional: GitHub, Stack Overflow, Kaggle
Social: Reddit, Hacker News
News: GDELT
Blogs: Medium, Towards Data Science

Limitations

Data is limited to the specified time range
Some sources may have rate limits or API restrictions
Citation counts may vary between sources
ML subfield classification is based on keyword matching and may not be perfect

Source

Generated by Research-Collector, an educational multi-source research aggregation tool.

License

MIT License

Citation

If you use this dataset, please cite the repository URL: https://huggingface.co/datasets/nellaivijay/llm-research-daily

Downloads last month: 196

Number of rows:

Total file size:

79.7 kB