hugging face releases 207 webgpu kernels for faster browser ai
hugging face launched a library of 207 optimized webgpu kernels and a browser benchmarking tool to speed up local ai inference.
machine learningtag tools
summaries filed under tools
open asr leaderboard adds hindi and indian english test sets
hugging face and voice arena release monsoon, a new open evaluation set for hindi and indian english speech recognition with detailed speaker metadata.
machine learninggemini omni 1.1 flash adds scene extension and 4k upscaling
google deepmind releases gemini omni 1.1 flash, a production-ready video model with scene extension, keyframe interpolation, 360p drafts, and 4k upscaling.
researchclaude code auto mode bypassed by prompt injection
researcher johann rehberger found a prompt injection attack that bypasses claude code's auto mode safety classifier 80% of the time.
tools4-bit model beats its full-precision source after quantization-aware healing
a new healing method lets a compressed, 4-bit model outperform its bfloat16 original on most benchmarks.
machine learninggranite 4.2: how ibm built its reasoning llms
ibm released granite 4.2, a family of dense reasoning llms in 3b, 8b, and 30b sizes, trained from scratch on 15t tokens with a multi-stage rl pipeline.
machine learningmemory-primed agent spawning for personal ai infrastructure
a system that spawns coding agents with pre-loaded memories from personal databases to avoid empty context windows.
researchConstraint-Aware GPU Allocator Boosts Utilization by 33 Points
A new GPU scheduler from Dharma-AI improves cluster utilization and priority-weighted output by planning allocations across the entire scheduling horizon instead of using FIFO order.
machine learningBenchmark optimization found in top speech recognition models
New probes show leading ASR models reproduce benchmark transcripts even when audio contradicts them, inflating scores.
machine learningAgent memory dosage depends on model capability
IBM Research shows that the right amount of agentic memory varies by model tier, with curated retrieval boosting weaker models and full guideline sets helping stronger ones.
machine learningLiquid AI Releases DSpark Draft Models for Faster LFM2.5 Inference
Liquid AI published DSpark draft checkpoints for three LFM2.5 models, enabling speculative decoding that speeds up inference up to 3.2x on GPU and 2.87x on-device without changing output quality.
machine learningLiquid AI Releases QAD Q4_0 Checkpoints for LFM2.5 Models
Liquid AI releases quantization-aware distillation checkpoints that recover 97% of BF16 accuracy for LFM2.5 models at Q4_0 memory and speed.
machine learningMojo programming language goes open source
Mojo, a Python-inspired language for GPU programming, has released its compiler and toolchain under Apache 2 license.
toolsSentence Transformers v6.0 Adds Multi-Vector Encoder
Sentence Transformers v6.0 introduces MultiVectorEncoder for ColBERT-style late interaction retrieval, supporting PyLate, Stanford-NLP ColBERT, and ColPali checkpoints.
machine learningFP8 Training on AMD GPUs Now Upstreamed in PyTorch
AMD and Meta engineers upstreamed FP8 training optimizations for AMD Instinct GPUs into TorchAO and TorchTitan, delivering up to 13.4% throughput gains on dense models and recovering 89% of quantization overhead on MoE models.
machine learningOpen models shift toward China, small models, and agents
Hugging Face data from early 2026 shows Chinese labs dominating frontier open models, small models driving downloads, and coding agents becoming a major user base.
machine learningWhat a 2,200-paper ICML reproduction effort found
A community hackathon used coding agents to reproduce over 2,200 ICML 2026 papers, finding that 23% had at least one falsified or contested claim.
machine learningGoogle DeepMind Launches Gemini 3.7 Flash
Google DeepMind introduces Gemini 3.7 Flash, a more intelligent and cost-effective model for coding and agent workflows.
researchMeta releases Muse Glimmer for on-device agentic AI via ExecuTorch
Meta open-sourced Muse Glimmer, a 30B-parameter model distilled from Muse Spark, optimized for on-device agentic workflows with ExecuTorch support for NVIDIA GPUs and Apple Silicon.
machine learningMeta releases Muse Glimmer, a 30B open-source multimodal model for local agents
Meta’s Muse Glimmer is a 30B-parameter multimodal model optimized for local, privacy-aware agentic tasks like coding and document analysis, released under Apache 2.0 with day-0 support in transformers and llama.cpp.
machine learning