Update README.md

README.md

@@ -12,107 +12,33 @@ pinned: false
 This repository hosts the official models and datasets released as part of the Human Behavior Atlas (HBA) project — a benchmark and foundation model ecosystem for unified multimodal social behavior understanding and reasoning.
 
 The resources include:
-- Unified behavioral foundation models
+- Unified OmniSapiens behavioral foundation models
 - Specialized downstream behavioral adapters
-- The Human Behavior Atlas benchmark datasets
+- The Human Behavior Atlas benchmark
 
 ---
 
-# Models
-
-## [OmniSapiens 2.0](https://huggingface.co/HumanBehaviorAtlas/OmniSapiens2.0)
-
-**Repository:** `HumanBehaviorAtlas/OmniSapiens2.0`
-
-**OmniSapiens 2.0** is the latest and strongest OmniSapiens foundation model for multimodal social behavior reasoning. The model is trained using **Heterogeneity-Aware Relative Policy Optimization (HARPO)** on the Human Behavior Atlas benchmark, enabling more balanced and consistent learning across diverse behavioral tasks and domains.
-
-The model supports unified reasoning over:
-- Emotion understanding
-- Humor detection
-- Sarcasm understanding
-- Sentiment analysis
-- Social reasoning
-- Behavioral interpretation
-
-This model represents the current state-of-the-art OmniSapiens release.
-
----
-
-## [OmniSapiens SFT](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_sft)
-
-**Repository:** `HumanBehaviorAtlas/omnisapiens_sft`
-
-**OmniSapiens SFT** is the supervised fine-tuned version of OmniSapiens trained on Human Behavior Atlas tasks using classification and QA-style supervision.
-
-This model serves as:
-- The supervised baseline prior to reasoning-RL training
-- The initialization model for subsequent RL-based OmniSapiens variants
-- A unified multimodal behavioral understanding model
-
----
-
-## [BAM — Humour Detection Adapter](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_bam_humour_detection)
-
-**Repository:** `HumanBehaviorAtlas/omnisapiens_bam_humour_detection`
-
-A specialized **Behavioral Adapter Module (BAM)** trained for multimodal humor detection tasks.
-
-This adapter is designed for downstream humor understanding applications while retaining the broader OmniSapiens behavioral representations.
-
 ---
 
-## [BAM — Sarcasm Detection Adapter](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_bam_sarcasm_detection)
-
-**Repository:** `HumanBehaviorAtlas/omnisapiens_bam_sarcasm_detection`
-
-A specialized **Behavioral Adapter Module (BAM)** optimized for sarcasm detection and sarcastic intent understanding across multimodal behavioral inputs.
-
----
-
-## [BAM — Sentiment Polarity (MOSEI) Adapter](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_bam_sentiment_polarity_mosei)
-
-**Repository:** `HumanBehaviorAtlas/omnisapiens_bam_sentiment_polarity_mosei`
-
-A specialized **Behavioral Adapter Module (BAM)** trained for sentiment polarity prediction on the MOSEI benchmark.
-
-The adapter supports multimodal sentiment understanding using behavioral and expressive cues.
-
----
-
-# Datasets
-
-## [Human Behavior Atlas (Parquet Version)](https://huggingface.co/datasets/HumanBehaviorAtlas/human_behavior_atlas)
-
-**Repository:** `HumanBehaviorAtlas/human_behavior_atlas`
-
-The primary release of the **Human Behavior Atlas (HBA)** benchmark.
-
-This version is distributed in **Parquet format**, which is generally easier and faster to use with the Hugging Face `datasets` library and most downstream training pipelines.
+# Models
 
-The benchmark contains diverse multimodal behavioral tasks spanning:
-- Emotion recognition
-- Humor understanding
-- Sarcasm detection
-- Sentiment analysis
-- Intent reasoning
-- Social reasoning
-- Mental health related behavioral understanding
+| Model | Description |
+|---|---|
+| [OmniSapiens 2.0](https://huggingface.co/HumanBehaviorAtlas/OmniSapiens2.0) | Latest and strongest OmniSapiens reasoning foundation model trained with HARPO for unified multimodal social behavior understanding. |
+| [OmniSapiens SFT](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_sft) | Supervised fine-tuned OmniSapiens model used as the initialization model for RL training. |
+| [BAM — Humour Detection](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_bam_humour_detection) | Behavioral Adapter Module specialized for multimodal humor detection. |
+| [BAM — Sarcasm Detection](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_bam_sarcasm_detection) | Behavioral Adapter Module specialized for sarcasm understanding. |
+| [BAM — Sentiment Polarity (MOSEI)](https://huggingface.co/HumanBehaviorAtlas/omnisapiens_bam_sentiment_polarity_mosei) | Behavioral Adapter Module for multimodal sentiment polarity prediction. |
 
+**OmniSapiens 2.0** is the latest and strongest OmniSapiens foundation model for multimodal social behavior reasoning. The model is trained using **Heterogeneity-Aware Relative Policy Optimization (HARPO)** on the Human Behavior Atlas benchmark, enabling more balanced and consistent learning across diverse behavioral tasks and domains. The model demonstrates strong multi-task performance, out-of-distribution generalization, and more interpretable reasoning traces.
 
 ---
 
-## [Human Behavior Atlas (TAR Version)](https://huggingface.co/datasets/HumanBehaviorAtlas/human_behavior_atlas_tar)
-
-**Repository:** `HumanBehaviorAtlas/human_behavior_atlas_tar`
-
-An alternative release of the Human Behavior Atlas benchmark distributed in **TAR file format**.
-
-This version may be useful for:
-- Custom preprocessing pipelines
-- Streaming-based workflows
-- Manual media extraction
-- Large-scale storage systems
+# Benchmark
 
-Most users may find the Parquet release (`human_behavior_atlas`) easier to download and integrate directly into training workflows.
+| Dataset | Description |
+|---|---|
+| [Human Behavior Atlas (Parquet)](https://huggingface.co/datasets/HumanBehaviorAtlas/human_behavior_atlas) | Primary HBA release optimized for Hugging Face `datasets` workflows. |
+| [Human Behavior Atlas (TAR)](https://huggingface.co/datasets/HumanBehaviorAtlas/human_behavior_atlas_tar) | TAR-based release for custom preprocessing and large-scale storage workflows. |
 
----
+The version distributed in **Parquet format**, is generally easier and faster to use with the Hugging Face `datasets` library and most downstream training pipelines. Although, we also provide the tar version to allow for flexible use.