Latest stories
StableVQ Targets the Hidden Instability in Vector-Quantized Image Tokenizers
A new Hugging Face paper argues that instability in vector-quantized tokenizers stems from entangled encoder-decoder and codebook...
Hugging Face Survey Maps LLM Evolution in Mental Health Care
A new survey published on Hugging Face organizes the fragmented literature on large language models in mental health into three ev...
Hugging Face Paper Introduces Bellman Policy Optimization, a Critic-Free RL Method for LLM Reasoning
A new paper on Hugging Face presents Bellman Policy Optimization (BPO), a critic-free method derived from Policy Mirror Descent an...
HyperQ Adds Quantum Circuit Branches to Frozen Diffusion Language Models
A new method, HyperQ, attaches token-conditioned quantum residual branches to a frozen masked-diffusion language model, training o...
GAE: A Geometry-Native Latent Space for 3D-Consistent World Generation
Researchers introduce the Geometry-Native Autoencoder (GAE), which reparameterizes geometry foundation model features into a compa...
RULER Turns Rubrics Into Rewards for Better SVG Generation
A new Hugging Face paper proposes rubric-based scoring as a more faithful signal for evaluating and optimizing SVG code generation...
OpenAI Enhances Prompt Caching for GPT-6
OpenAI says GPT-6 introduces improved prompt caching with higher cache hit rates, new diagnostics, explicit breakpoints, and contr...
OpenAI Unveils GPT-6 Sol and Luna, Two New Frontier Models
OpenAI has introduced GPT-6 Sol and Luna, two models designed to bring frontier intelligence to everyday work with different balan...
Anthropic Launches Claude Opus 5.5: Frontier Performance at 40% Lower Cost
Anthropic unveils Claude Opus 5.5, the first model in its new Claude 5.5 family, matching Claude Fable 5.1's performance on most w...
Parallel Halves Research Time and Cost with OpenAI's GPT-6 Astra
OpenAI says Parallel's agents used GPT-6 Astra to research and synthesize labor-market data in half the time and at half the cost...
HuRo Turns Human Videos into Robot Training Data, Boosting Manipulation Success to 80.3%
A new Hugging Face paper introduces HuRo, a robotization pipeline that converts heterogeneous human videos into robot-aligned obse...
Hugging Face Paper Traces Video Diffusion Physics Flaws to RoPE Attention Decay
A new interpretability study on Hugging Face examines how text-to-video diffusion models plan motion, finding that Rotary Position...