Skip to content
K = 0.731
All news · Today8 items

#Reasoning

All collected and published items, ordered by time. Picks are a selected subset.
6/11Thu
  1. Google DeepMind71

    Google DeepMind releases DiffusionGemma, generating text up to 4× faster

    Google DeepMind has released DiffusionGemma, an experimental open-source model that uses text diffusion to generate entire chunks of text in parallel. It generates text up to 4× faster on specialized GPUs, exceeding 1,000 tokens per second on a single NVIDIA H100 and 700 tokens per second on an RTX 5090.

    Why it matters: Google DeepMind's open text-diffusion model comes with concrete speed and hardware figures, helping readers assess the feasibility of local interactive inference.

6/10Wed
  1. Andrej Karpathy · X78

    Karpathy on Claude Fable 5: same base model as Mythos, with a major leap in long-horizon tasks

    Andrej Karpathy says Claude Fable 5 shares the same underlying model as Mythos but adds safety safeguards, and reaches state-of-the-art results by some margin on almost every benchmark.

    Why it matters: Drawing on firsthand use, Karpathy calls Fable 5 a release worthy of a major version-number jump and describes concrete changes on long-horizon tasks.

6/9Tue
  1. Google DeepMind78

    Google DeepMind releases Gemma 4 12B, a unified encoder-free multimodal model

    Google DeepMind has released Gemma 4 12B, an agentic multimodal model aimed at laptops, positioned between the edge-focused E4B and the 26B MoE. It is the company's first medium-sized model with native audio input.

    Why it matters: Gemma 4 12B removes multimodal encoders and feeds visual and audio input directly into the LLM, giving readers a view of hardware requirements for local multimodal agents.

6/8Mon
  1. Import AI62

    Import AI 460: social rules can be reward-hacked, Anthropic's RSI data, and RL drone racing

    This issue of Import AI summarizes three studies. Researchers from King's College London, Fudan University, and the Alan Turing Institute built SocioHack, a benchmark with 72 sandboxed social environments testing whether AI can game rules in scenarios such as credit-card rewards and school grading. LLMs trained with RL rediscovered patched vulnerabilities in historical environments with 61.25% recall and 90.85% precision.

5/29Fri
5/8Fri
5/4Mon
4/22Wed
  1. Google Research62

    Google proposes ReasoningBank, helping agents learn from successes and failures

    In an ICLR paper, Google Research proposes ReasoningBank, an agent-memory framework that extracts high-level, structured reasoning memories from successful and failed experiences. Its code is open-source.

    Why it matters: The paper describes a closed loop for extracting reasoning memories from successes and failures, with success-rate and step-count results on WebArena and SWE-Bench-Verified.

4/20Mon
  1. Berkeley AI Research36

    GRASP: gradient-based long-horizon planning for world models

    A Berkeley AI Research team has proposed GRASP, a gradient-based planner that makes long-horizon planning with learned dynamics (world models) feasible. It lifts trajectories into virtual states for parallel optimization across time, injects stochasticity directly during iterative state refinement to explore, and reshapes gradients so actions receive clearer signals. The method addresses ill-conditioned computation graphs, local minima caused by non-greedy structure, and fragile gradients in high-dimensional visual models.

4/16Thu
3/25Wed
  1. Google Research62

    Google releases TurboQuant vector-compression algorithm, reducing KV cache to 3 bits without loss of accuracy

    Google has released TurboQuant, a compression algorithm that quantizes KV cache to 3 bits without training or fine-tuning and without loss of model accuracy, reducing KV-memory use by at least 6×.

    Why it matters: Google reports TurboQuant's algorithm details and benchmark results, helping readers assess the usefulness of KV-cache compression for long-context workloads.

3/17Tue
  1. Mistral AI78

    Mistral releases Mistral Small 4, unifying reasoning, multimodal, and coding capabilities

    Mistral AI has released Mistral Small 4, the next major version of the Mistral Small family and its first model to combine Magistral's reasoning, Pixtral's multimodal, and Devstral's agentic-coding capabilities in one model. It is licensed under Apache 2.0.

    Why it matters: Mistral has combined reasoning, multimodal, and coding capabilities in one open small model, with adjustable reasoning effort and deployment requirements to help assess implementation costs.

3/13Fri
  1. Berkeley AI Research38

    Berkeley AI Research proposes SPEX and ProxySPEX for identifying key LLM interactions at scale

    Berkeley AI Research has proposed SPEX and ProxySPEX, algorithms that identify important interactions driving LLM outputs across feature, data, and model-component attribution. SPEX turns interaction search into a sparse-recovery problem by exploiting sparsity and low-order structure; ProxySPEX uses additional hierarchy to match SPEX performance with about ten times fewer ablation experiments.

11/1Sat
  1. Berkeley AI Research36

    Reinforcement learning without TD: scaling off-policy RL with divide and conquer

    Berkeley AI Research has proposed a divide-and-conquer off-policy reinforcement-learning algorithm that does not rely on temporal-difference (TD) learning. It reduces the number of Bellman-recursion steps from linear to logarithmic, enabling scaling to long-horizon tasks. The method achieves scaling for the first time in goal-conditioned RL, using the triangle inequality for shortest-path distances between states to construct transitive Bellman updates. Selecting the optimal subgoal w remains an open problem.

7/17Thu
  1. Mistral AI62

    Mistral adds Deep Research, voice mode, and more to Le Chat

    Mistral has introduced several new features in Le Chat, including a preview of Deep Research mode, voice mode powered by the new Voxtral model, multilingual reasoning supported by the Magistral reasoning model, Projects for organizing conversations, and image editing in partnership with Black Forest Labs.

    Why it matters: Mistral has brought deep research, voice, reasoning, and image editing into Le Chat at once, giving readers a view of the assistant's expanding capabilities.

6/10Tue