Updated
Updated
Items with an AI score under 20 are hidden. Show low-relevance items
AIHamel Husain and Shreya Shankar published evals skills, a set of skills for AI product evals that helps users avoid common mistakes. The entry point evals-start routes users to eval-audit for existing pipelines or error-discovery for unanalyzed traces. The repository is available at ai-evals-course/evals-skills and installs via npx skills add.
Jim Fan@DrJimFanAI score75AIJim Fan announced DreamDojo, an open-source interactive world model that takes robot motor controls and generates future frames in pixels. It is pre-trained on 44K hours of human egocentric video using latent actions, then post-trained onto specific robot hardware, and a real-time version runs at 10 FPS for live teleoperation, policy evaluation, and model-based planning. The author reports a +17% real-world success gain on a fruit packing task, and weights, code, datasets, and the whitepaper are released.
Guillaume Lample @ NeurIPS 2024@GuillaumeLampleAI score48AIMistral has released the Voxtral-Mini-4B-Realtime-2602 model weights on Hugging Face, alongside an arXiv paper and a live realtime audio playground in Mistral AI Studio. The post links these resources but provides no further benchmark figures or capability details.
Guillaume Lample @ NeurIPS 2024@GuillaumeLampleAI score62AIMistral announces Mini Transcribe 2 via API at $0.003 per minute and a realtime transcription option at $0.006 per minute. The realtime model's open weights are published on Hugging Face, alongside a realtime demo and a blog post.
Guillaume Lample @ NeurIPS 2024@GuillaumeLampleAI score62AIVoxtral Realtime is a natively streaming speech model for voice agents and live applications, with latency configurable down to sub-200ms. At 480ms it stays within 1-2% WER of the offline model, and the weights are released under Apache 2.0. The attached FLEURS chart compares word error rates across latency settings for ten languages, including Chinese.
Guillaume Lample @ NeurIPS 2024@GuillaumeLampleAI score62AIMistral announces Voxtral 2 with two transcription models: Voxtral Realtime, released under an Apache 2 license with latency configurable to sub-200 ms, and Voxtral Mini Transcribe 2, which adds speaker diarization, word-level timestamps, and context biasing. The models support 13 languages and are available through the Mistral API, which the post describes as one of the most cost-effective transcription APIs on the market. The attached chart shows word error rates on FLEURS across Italian, Spanish, English, German, Portuguese, French, Russian, Dutch, and Chinese at several latency settings.
AIAi2's Open Coding Agents family, with SERA as its first release, was built by Tim Dettmers and collaborators on 32 GPUs. The method generates synthetic bug trajectories with soft verification, comparing patches by line overlap instead of running tests. The post reports that a 32B model fine-tuned on about 7,000 trajectories for one private repository matched its GLM 4.5-Air teacher, and that the baseline costs $500 to run.
BAAI@BAAIBeijingAI score40AIBAAI released RoboBrain 2.5, an embodied AI model that turns 2D scene understanding into actionable 3D trajectories and provides dense temporal value estimates for real-time progress feedback on long-horizon tasks. The post says it achieves SOTA across multiple spatial and temporal reasoning benchmarks, though it names no specific scores. Project page, paper, GitHub code, and model weights are linked.
BAAI@BAAIBeijingAI score38AIBAAI's RoboCOIN is a bimanual robot dataset with more than 180,000 trajectories across 421 tasks, collected from 15 robot platforms in 16 real-world scenarios. Its three-tier annotations at trajectory, segment, and frame levels help robots learn both what to do and how to do it. The post says integrating these annotations raised success rates on complex tasks by up to 50% for models such as π₀.
AIBlack Forest Labs has published FLUX.2 [klein] 4B Base, a 4 billion parameter text-to-image model that also supports multi-reference editing. The model is undistilled, is released with open weights under Apache 2.0, and is described as fitting in about 13GB VRAM on cards such as the RTX 3090 or 4070, with reference code available in its GitHub repository and support in ComfyUI and Diffusers.
AIBlack Forest Labs has released FLUX.2 [klein] 9B Base, a 9 billion parameter undistilled rectified flow transformer with open weights for text-to-image generation and multi-reference editing. The model is intended for fine-tuning, LoRA training, and research, and fits in about 29GB VRAM on NVIDIA RTX 4090-class GPUs. A reference implementation is available on GitHub, and the model works with ComfyUI and Diffusers.
AIFunAudioLLM has released Fun-Audio-Chat-8B, a roughly 8B-parameter large audio language model for natural, low-latency voice interaction, under Apache 2.0. It uses Dual-Resolution Speech Representations with a 5Hz frame rate, which the source says reduces GPU hours by nearly 50%, and it supports English and Chinese.
AIAlibaba's FunAudioLLM released Fun-ASR-MLT-Nano-2512, an 800M-parameter multilingual speech recognition checkpoint on Hugging Face that supports 31 languages, with emphasis on East and Southeast Asian languages. It is trained on hundreds of thousands of hours of speech and is available through the FunASR toolkit. The source's benchmark tables cover the Fun-ASR family rather than this checkpoint, so no checkpoint-specific accuracy figures are reported.
AITongyi Lab has released Fun-ASR-Nano-2512, an end-to-end speech recognition large model trained on tens of millions of hours of real speech, supporting low-latency real-time transcription across 31 languages. The model, which has 800M parameters, targets industry use such as education and finance and claims 93% accuracy in far-field, high-noise conditions. It is available on Hugging Face and works with the FunASR toolkit.
Mira Murati@miramuratiAI score46AIThinking Machines has made Tinker generally available, adding support for advanced vision input models and Kimi K2 Thinking. The update also includes a simpler way to sample from models.
AIAlibaba's FunAudioLLM has released Fun-CosyVoice3-0.5B-2512, a 0.5B-parameter LLM-based text-to-speech model on Hugging Face, with an RL variant also published. The model supports zero-shot voice cloning across 9 languages and 18+ Chinese dialects and accents, with streaming output at latency as low as 150ms. On the source's test-en benchmark, it reports a 2.24% WER and 71.8% speaker similarity, and the RL version reports 1.68% WER.
AIAndrej Karpathy built hn-time-capsule, a tool that feeds each December 2015 Hacker News front-page article and its comment thread to GPT-5.1 Thinking for a retrospective analysis. The project, written with Claude Opus 4.5 in about three hours, processes 930 articles at a cost of about $58 and roughly one hour. Results include prescience and wrongness grades for commenters, and the project is hosted on his website with the intermediate data available for download.
AIARC Prize reports that the top Kaggle entry reached 24% on the ARC-AGI-2 private dataset at $0.20 per task, and that all winning solutions and papers are open source. The top verified commercial model, Opus 4.5 (Thinking, 64k), scored 37.6% at $2.20 per task, while a Poetiq refinement on Gemini 3 Pro reached 54% at $30 per task. The author argues that refinement loops are the main driver of 2025 progress, and says ARC-AGI-3 is planned for early 2026.
Why it matters: The post links 2025 competition results to a broader argument about refinement loops, showing how benchmark outcomes are being read as evidence of AI reasoning progress.
AIApple's CLaRa-7B-E2E is a fully end-to-end unified RAG model that jointly optimizes retrieval and generation, with 16x and 128x document compression. It is trained with end-to-end finetuning using differentiable top-k retrieval and a unified language-modeling objective. The model is available on Hugging Face with example end-to-end inference code.
AIMoonshot AI has released Kimi-Linear-Base and Kimi-Linear-Instruct, both 48B total and 3B activated parameters with a 1M context length, on Hugging Face. The models use Kimi Delta Attention in a 3:1 hybrid ratio with global MLA, cutting KV cache by up to 75% and boosting decoding throughput by up to 6x at 1M tokens. The KDA kernel is open-sourced in FLA, and the checkpoints were trained on 5.7T tokens.
Why it matters: The model card gives concrete throughput and KV cache figures for a hybrid attention design, which helps readers weigh its long-context tradeoffs against full attention.
AICognition is bringing back its Devin Open Source Initiative, offering $500 in Devin ACU credits to open-source GitHub projects with over 100 forks. Projects below that threshold will still be considered. Eligible projects must have an OSI-approved license and be actively maintained, and maintainers can apply through a linked form.
AICrossmint said Devin became its top contributor to the open-source GOAT SDK during an initial trial, merging 8 pull requests versus 4 for the next contributor. Examples included a DEXScreener plugin built from a documentation URL and a harder Sui blockchain integration that needed three rounds of feedback and about an hour of human involvement. The company said its results depended on proper training, clear task context, and planned validation, not on treating Devin as superhuman.
AICognition is launching the Devin Open Source Initiative, giving selected open source maintainers 500 free ACUs on a Devin Teams plan as part of Devin's general availability launch. The post shows Devin contributing pull requests to projects including Anthropic's MCP Inspector, Dagger, and nanoGPT, with maintainers still reviewing the results. Devin's GitHub integration forwards PR comments and CI checks to help refine changes, though the company warns a human should still verify final quality.
AILiquid AI reports STAR, an evolutionary algorithm that synthesizes tailored neural network architectures from a numerical genome representation. The authors say it produced hundreds of designs that outperform Transformer and hybrid architectures in quality, with smaller caches and parameter counts, and can optimize for latency on target hardware. The full method is described in the arXiv technical report 2411.17800.
Why it matters: The post explains how evolutionary search over a new architecture design space produced designs beating Transformers and hybrids, giving a concrete method for quality versus latency and memory trade-offs.