Favorites
Collection
1 item • Updated • 1
asisai
asis-ai
AI & ML interests
Image/video automation, generative AI production workflows
Recent Activity
upvoted a collection about 12 hours ago
Favorites upvoted an article about 12 hours ago
This Title Is Already Tokenized (Tokun P.2)Organizations
upvoted a collection about 12 hours ago
upvoted an article about 12 hours ago
Article
This Title Is Already Tokenized (Tokun P.2)
apehex
• • 6 reacted to John6666's post with 👍 about 12 hours ago
Post
40484
I used up my Zero GPU Quota yesterday (about 12 hours ago). At the time, I got a message saying “Retry at 13:45 (approx.)”, but now it's just changed to “Retry at 03:22”.
Anyway, everyone, let's be careful not to use up our Quota...
Related: https://huggingface.co/posts/Keltezaa/754755723533287#67e6ed5e3394f1ed9ca41dbd
Anyway, everyone, let's be careful not to use up our Quota...
Related: https://huggingface.co/posts/Keltezaa/754755723533287#67e6ed5e3394f1ed9ca41dbd
- 1 reply
upvoted an article about 12 hours ago
Article
📄 Research Papers Dataset
tegridydev
• • 3 upvoted a paper about 12 hours ago
upvoted a collection about 12 hours ago
upvoted a paper about 12 hours ago
upvoted an article about 12 hours ago
Article
MIEB: The Benchmark That Stress-Tests Image-Text Embeddings Like Never Before
isaacchung
• • 18 upvoted a paper 2 days ago
reacted to pbhappliedsystems's post with 👀 2 days ago
Post
2278
🚀 **New flagship dataset — and an argument about what a dataset card should be.**
Most synthetic datasets on the Hub ship row counts, a license, and little else — pipeline opaque, rejection criteria unstated, compliance unaudited. We published the opposite.
**SynthEval Cloud — Regulated-Domain Synthetic Instruction Dataset**
👉 pbhappliedsystems/syntheval-cloud-regulated-instruct-1k
**1,116** quality-gated instruction records across **7 regulated domains** (medical, legal, GDPR, privacy, education, e-commerce, transport). Every record cleared a documented cascade, not a vibe check:
- 🧪 **Dual-signal hallucination gate** — rejects only when embedding cosine *and* keyword-overlap both fail; a low score alone never rejects.
- 🔒 **Layered PII masking + independent leak audit** — a separate over-reporting scanner found **0.0% residual leak** across all 1,116 records.
- 📊 **Whole-corpus evaluation, not a sample** — MATTR **0.769**, mean cosine **0.73**, **0%** near-duplicates, **96.9%** yield.
- 🧾 **The 36 rejections ship too**, each tagged with its failing gate. Removal at the gate is the product; we show our work.
Every number on the card is a field in the
One release from **SynthEval**: Studio (local GPU) + Cloud (Modal+vLLM), proving quality parity across substrates.
📄 Whitepaper: https://pbhappliedsystems.com/SynthEval_Studio_and_Cloud_Quality-Gated_Synthetic_Data_Generation.pdf
🔎 Overview: https://pbhappliedsystems.com/synthetic-data.html
**CC BY 4.0** — commercial use welcome, just credit it. Need defensible synthetic data at scale? Let's talk.
— Patrick Hill, PBH Applied Systems
Most synthetic datasets on the Hub ship row counts, a license, and little else — pipeline opaque, rejection criteria unstated, compliance unaudited. We published the opposite.
**SynthEval Cloud — Regulated-Domain Synthetic Instruction Dataset**
👉 pbhappliedsystems/syntheval-cloud-regulated-instruct-1k
**1,116** quality-gated instruction records across **7 regulated domains** (medical, legal, GDPR, privacy, education, e-commerce, transport). Every record cleared a documented cascade, not a vibe check:
- 🧪 **Dual-signal hallucination gate** — rejects only when embedding cosine *and* keyword-overlap both fail; a low score alone never rejects.
- 🔒 **Layered PII masking + independent leak audit** — a separate over-reporting scanner found **0.0% residual leak** across all 1,116 records.
- 📊 **Whole-corpus evaluation, not a sample** — MATTR **0.769**, mean cosine **0.73**, **0%** near-duplicates, **96.9%** yield.
- 🧾 **The 36 rejections ship too**, each tagged with its failing gate. Removal at the gate is the product; we show our work.
Every number on the card is a field in the
evaluation_report.json shipped beside the data — full methodology + provenance (Mistral-Nemo AWQ W4A16 · vLLM 0.8.5.post1 · Modal A10G).One release from **SynthEval**: Studio (local GPU) + Cloud (Modal+vLLM), proving quality parity across substrates.
📄 Whitepaper: https://pbhappliedsystems.com/SynthEval_Studio_and_Cloud_Quality-Gated_Synthetic_Data_Generation.pdf
🔎 Overview: https://pbhappliedsystems.com/synthetic-data.html
**CC BY 4.0** — commercial use welcome, just credit it. Need defensible synthetic data at scale? Let's talk.
— Patrick Hill, PBH Applied Systems
updated a
Space 2 days ago
published 2
Spaces 2 days ago
updated a
Space 3 days ago