AI News | UncensoredHub

Loading…

This week in open AI: on-device inference meets agentic breakthroughs

Jun 2, 2026

Two trends collided this week: pushing frontier-scale models onto consumer hardware and teaching agents to navigate messier, real-world environments. Nvidia and Liquid AI both announced architectures that fit double-digit-billion-parameter models into phone and laptop memory, while researchers demonstrated that reinforcement learning on live websites and fairness-aware policy synthesis can unlock new agent capabilities. Meanwhile, the open fine-tune ecosystem continued its rapid expansion.

In this digest

OpenWebRL-4B reaches 67% success on live-web tasks with minimal supervised data — Training web agents on real sites with only 400 seed examples just hit 67% success—online RL may finally beat the sim-to-real gap.
MobileMoE scales sparse experts to smartphones, cuts inference by 2–4× — Sparse experts cut smartphone inference time in half without ballooning memory—MoE design is now a mobile-first architecture question.
Autoresearch agent discovers fairness mechanisms emerge only under maximin objectives — An autonomous researcher agent just proved fairness mechanisms only emerge when you optimize for equity, not efficiency—a result no human study isolated.
Liquid AI's LFM2.5-8B-A1B runs on phones with 1.5B active parameters — Liquid AI's 8B-parameter MoE activates only 1.5B per pass and ships for phones, laptops, and robots—on-device just got a new baseline.
ProRL corrects gradient bias in recommendation path planning — ProRL fixes two gradient-bias bugs that have plagued proactive recommenders, beating baselines on three real datasets—RL debugging pays off.
Claude Opus 4.8 catches 4× more code bugs, leads on browser-agent benchmarks — Claude Opus 4.8 catches four times more code bugs and leads browser-agent benchmarks—Anthropic is betting on reliability over raw scale.
Nvidia RTX Spark Superchip brings 120B local LLMs to Windows PCs — Nvidia's RTX Spark Superchip pairs Arm cores with 128 GB unified RAM to run 120B models in FP4 on Windows PCs—local inference just leapfrogged a generation.
Illustrious XL 1.0 Improved Uncensored v3.0 SDXL fine-tune lands on HuggingFace — Illustrious XL 1.0 Improved Uncensored v3.0 lands on HuggingFace with safetensors—anime-focused SDXL fine-tunes remain a prolific niche.

Also this week

Anthropic files confidential S-1 with SEC, moving toward IPO — Anthropic filed a confidential S-1 with the SEC, the first formal step toward a public offering; timing and valuation remain undisclosed.
BitsMoE cuts Qwen3-30B 2-bit quantization time by 12× with spectral bit allocation — BitsMoE uses SVD-based spectral bit allocation to quantize Qwen3-30B to 2 bits 12× faster, gaining 27.83 points over GPTQ.
Nova Anime AM hits 8,000 downloads with optimized low-step sampling — Nova Anime AM, an Anima checkpoint optimized for low-step anime generation, hit 8,060 downloads on Civitai.
Google used Gemini to produce I/O 2026 conference — Google disclosed it used Gemini to handle production workflows for I/O 2026, the first public confirmation of AI staging the flagship conference.
Qwen-Image NSFW LoRA launches on HuggingFace with Diffusers support — Gigo1990 released an uncensored LoRA for Alibaba's Qwen-Image on HuggingFace, enabling local text-to-image without safety filters.
osmKeye-VL-2.0-30B abliterated weights land in three quantizations on HuggingFace — osmapi published abliterated weights for osmKeye-VL-2.0-30B-A3B in full precision and two MLX quantizations on HuggingFace.

Open the digest →

What's happening

Anatomical slider LoRA extends Illustrious XL and Pony model control

Gemma 4 26B uncensored weights quantized to GGUF for local inference

Microsoft's MAI-Thinking-1: 35B-active MoE model trained on 8,000 GB200s

Google DeepMind open-sources Science Skills for AI research workflows

Anthropic adds Services Track and Partner Hub to Claude deployment network

Ideogram 4 open-weights: 9.3B DiT with Qwen vision encoder, native 2K

q0 framework trades compute for data via parallel ensembles and cyclic schedules

LeJEPA latents recover world state up to rotation, LeCun team proves

Stereological theory reveals LLM benchmark blind spots 100× larger than score gaps

MisoTTS 8B clones voices in 110ms with 10-second audio samples

Ideogram-4 open weights challenge FLUX and Hunyuan at 9.3B parameters

NVIDIA Nemotron 3 Ultra: 550B sparse model activates 55B weights for agent tasks

Claude writes 80% of Anthropic's production code; agents outpace humans on research tasks

Cognition's $10M productivity guarantee bets on Devin's measurable value

MAOAM unifies object and material selection via text or click

ChatGPT Dreaming synthesizes memory from chat history, cuts compute costs fivefold

Music recommender system lifts recall 95% by fusing lyrics, audio, and LLM metadata

Gemma 4 12B abliterated checkpoints surface on HuggingFace within 48 hours