AIREP v2 is now public on arXiv and indexed in Hugging Face Daily Papers.

The main change is a stricter evidence model for AI runtime governance:

Decision → Control → Execution → Effect

These are separate evidence stages — a decision does not prove delivery, delivery does not prove execution, and execution does not prove the intended effect.

v2 also defines bounded AIREP-Core / Authenticated / Witnessed assurance and structured reconciliation that preserves missing, unevaluated and indeterminate evidence instead of turning absence into success.

v0.2.0-beta.1 remains experimental. No same-version producer→consumer interoperability claim is made.

📄 Paper
AIREP: A Protocol for Per-Decision Evidence in AI Runtime Governance (2608.21363)

🔬 Runnable Evidence Lab
phionyx/airep-evidence-lab

💻 Code / spec / evidence
https://github.com/halvrenofviryel/ai-runtime-evidence-protocol

reacted to pavle-scalably's post with 😎 10 days ago

Post

2288

14 days serving

unsloth/Qwen3.8-27B-NVFP4 to production agents on 2x RTX 5090 (vLLM 0.27, TP=2, 262K context, FP8 KV):

28,097 requests, 860.6M prompt tokens, 82.6% prefix-cache hit rate, TTFT p50 0.61 s, 0 engine errors.

The observation: prefix cache, not throughput, decides whether a 27B model keeps up with agents.

Mean request is 30,100 tokens in, 983 out, because every turn resends the whole session. Two flags mattered most: --max-num-seqs 12 (queue p95 went 9.4 s to 233.6 s past that) and --watermark 0.08 (preemptions 29 to 2). And thinking off for tool loops: 917 tokens in 11.7 s vs 11,170 in 144 s, same answer.

Full config and counters: scalably.io/blog/qwen3-8-27b-nvfp4-rtx-5090-production

Next we are preparing an 8x B300 node in an EU data center for open-weight serving. Which models or workloads are underserved for you?

7 replies

reacted to snkii's post with 🤗 10 days ago

Post

841

Sori-1B-MCQ — a 1B audio-language model that answers your multiple-choice questions about what it hears: one of your options, a probability for each, and a confidence. Inspired by TypeSafe's System One model, Jev.

snkii/Sori-1B-MCQ

reacted to Banaxi-Tech's post with 👍 10 days ago

Post

109

We're introducing Layer-Feedback-Transformers (LFT)
Check out our article here: https://huggingface.co/blog/Banaxi-Tech/layer-feedback-transformer-lft

31 replies

reacted to SoulInPsyAbstract's post with 🤝 10 days ago

Post

1118

3 things this stop-gate didn't have last week:
* Rejected human-in-the-loop gate: Dropped the step-up human confirmation idea, recognizing that a capable agent will manipulate the human (as demonstrated in GPT-4's System Card where it hired a TaskRabbit worker) and that single-reviewer oversight quickly degrades into meaningless rubber-stamping.
* Verifier trust (Auth0 M2M signatures): Moved away from self-reported STOP verdicts by having verdicts signed using an RS256 JWT via a separate Auth0 M2M identity that the executing agent cannot access or fake internally.
* Separation of dangerous vs. malicious: Formalized distinct paths for risk—risky-but-honest actions trigger a blocked action while the agent keeps running, whereas a misbehavior-discriminator flagging deception in the trace quarantines the entire agent for subsequent human review.

reacted to ProCreations's post with 🤝 10 days ago

Post

1544

Introducing Bonsai 2 27b GSQ RCO! It applies two newly-popular methods for quants to retain higher accuracy. Bonsai 2 27b GSQ RCO achieves around 6 percent lower perplexity on WikiText-2 compared to Bonsai 2 27b and roughly unchanged benchmark accuracy overall, with small mixed differences. It stays under 7gb, staying small like the original bonsai. Note that this is more of an experiment than a true finished product but the gains we saw are cool! Test it out and let me know what you think!
ProCreations/bonsai-2-27b-gsq-rco-gguf

upvoted a changelog 10 days ago

Hugging Face Changelog

Granular Feature Access

Aug 12

• 404

New activity in huggingface/InferenceSupport 10 days ago

Mohammedkarimi/argon1

🤝 1

#12504 opened 10 days ago by

Saeed1361

New activity in Mohammedkarimi/argon1 11 days ago

سلام

👍 1

#1 opened 11 days ago by

Saeed1361

reacted to nwaughachukwuma's post with 🔥❤️ 12 days ago

Post

2725

Following this post - https://huggingface.co/posts/nwaughachukwuma/366831808712777, I made another run on a fresh RTX PRO 4000 using inkling, gemini-3.5-flash, and glm-5.2+mm-ctx

| case | gemini-3.5-flash @k=1 | glm-5.2+mm @k=1 | inkling @k=1 |
| --- | --- | --- | --- |
| **MEAN correctness** | **78.2** | **79.5** | **57.9** |
| **MEAN speed (s)** | **111** | **123** | **69** |
| **completed (case-runs)** | **20/23** | **21/23** | **18/23** |