NAll news

All news

Wednesday22 items

  3. Rohan PaulResearch
    AI score79

    Claude autonomously designed 17 molecules that bind scorpion venom neurotoxins in Claude Science

    In Anthropic's Claude Science environment, Claude autonomously completed an entire round of protein design as an agent: it selected the attack sites on scorpion venom neurotoxins on its own, drove 12 protein design programs, set scoring thresholds, and selected 180 molecules from the 318,484 generated to send for physical synthesis; of the 211 designs tested by Adaptyv Bio and Twist Bioscience, 17 successfully bound the target toxin. The request was made by Citrini Research, a financial firm with no laboratory or protein expertise. Anthropic did not call it an antivenom serum because the tests needed to support that claim have not yet been conducted.

    Quoted postAnthropic · @AnthropicAI

    New on our Science Blog: designing protein binders against scorpion neurotoxins. Scorpions sting about 1.2m people every year, and antivenoms are often costly or in short supply. Claude has made an early step toward finding new treatments. https://t.co/l6LgdEqM7I

    #English text is being prepared.#Agents#Anthropic

  5. Rohan PaulResearch
    AI score56

    Meta's new paper proposes the ProPerMem benchmark: proactive assistants struggle with autonomously judging when to interrupt

    A new Meta paper points out that the failure point of proactive assistants (such as AI glasses) lies in autonomously judging when to speak, not just what to say. The team built the ProPerMem benchmark, using first-person video paired with interference-rich, full-year memory averaging 618K tokens, and featuring human-verified trigger windows. All baselines scored at most 15%, even frontier models using retrieval-augmented memory; when oracle evidence was provided within the window, GPT-5.3-Chat reached 96%, but when it was left to choose the timing itself, this dropped to 38%, and to 15% after adding retrieval. Grasping both content and timing while reducing trigger frequency can raise accuracy from under 10% to 42.4%.

    #Papers/Research#Agents#Multimodal

  7. Jiwei.comResearch
    AI score58

    A team at Tsinghua University has developed Airscope, a 1-gram wireless mesoscopic microscope, with the results published in Nature

    The team led by Dai Qionghai of Tsinghua University and the team led by Dai Pu of Chinese PLA General Hospital have developed Airscope, a wireless mesoscopic microscope weighing only 1 gram. It can record neural activity in the mouse cerebral cortex within a 6-millimeter field of view at 4 to 6 micrometers single-cell resolution, and a single mouse can have more than 8,500 neurons recorded simultaneously, 25 times that of existing devices of the same type. The device uses only 4 aspheric lenses and achieves on-device compression through a 2-megapixel image sensor, enabling wireless transmission of megapixel-level images and freeing it from cable constraints. The results were published online in Nature on the 19th, and the team has used it to record cortical activity in multiple mice in a 5-square-meter natural habitat and a large maze with 32 endpoints.

    #Papers/Research#English text is being prepared.

  9. Ai2 BlogFirst-partyResearch
    AI score61

    Ai2 releases the FairStress benchmark to evaluate group fairness in LLM decision-making

    Ai2 introduces FairStress, a benchmark measuring the demographic group fairness of LLMs in decisions such as hiring, healthcare, and lending. It contains 300 two-candidate scenarios across 29 domains, and expands to more than 13 million test cases after combining identity presentation methods with adversarial prefixes. It is now available for download and integrated into Olmo-eval. The experiments cover the Olmo, Llama, Qwen, Gemma, Claude, and GPT series, and find that when candidate identities are presented implicitly, the average gap at which models abandon the correct answer is 6.9 percentage points, and 9.3 percentage points when identities are explicitly labeled. On FairStress-Core, Claude Opus 5 and GPT-5.5 select the minority-coded candidate about 60–65% of the time when there is no clearly superior candidate; most of the nine debiasing methods tested fail to reliably narrow the asymmetry, while counterfactual chain-of-thought narrows the gap while maintaining accuracy.

    #Evaluations / benchmarks#Safety / alignment#Open-source ecosystem

Tuesday28 items

  1. METR BlogFirst-partyResearch
    AI score65

    METR reviews the misalignment risk chapter of Anthropic's August 2026 risk report

    METR conducted an external review of the "Misalignment in high-risk scenarios" chapter in the August 2026 edition of Anthropic's risk report. It agreed that the catastrophic misalignment risk of the covered models (including Mythos 5 and the internal model Model 2) is low, but held that this is mainly due to limited model capabilities rather than safeguards and alignment training. METR pointed out that the covert capability benchmark in the report cannot represent a model's ability to evade Anthropic's internal monitoring systems, and stated that recent events weaken the report's argument that unknown severe situational misalignment is "unlikely." METR also recommended disclosing four redacted pieces of information and conducting end-to-end red-team testing of the monitoring system.

    #Safety / alignment#Evaluations / benchmarks#Anthropic

  2. Hugging FaceFirst-partyResearch
    AI score63

    ALTK-Evolve Uses Learned Memory to Put GPT-4.1 Customer-Service Agent Atop the τ²-bench Retail Leaderboard

    IBM Research has released ALTK-Evolve on Hugging Face. It uses GPT-4.1 to distill behavioral guidelines from the agent's own training trajectories and append them to the system prompt, achieving 67.5% Pass^4 on the τ²-bench retail task and surpassing all retail entries in its October 12, 2026 leaderboard snapshot. The method requires no weight updates or human annotation. Retail average accuracy rises from 78.1% to 82.5%, and Pass^4 rises from 57.5% to 67.5%; in the airline domain, Pass^4 rises from 35.0% to 45.0% with GPT-4.1 and to 82.0% with GPT-5.5. The authors say the gains come from more accurate decisions rather than longer conversations; the guidelines slightly reduce the number of dialogue turns, and the added cost is mainly in input tokens.

    Same story · 2 reports

    #Papers/Research#Agents#Evaluations / benchmarks

  5. AI at MetaFirst-partyResearch
    AI score71

    Meta releases Muse Glimmer technical report, a 30B open-weight model targeting agent tasks on consumer-grade hardware

    Meta releases the Muse Glimmer technical report, a 30B open-weight model targeting agent tasks on consumer-grade hardware. At 4-bit precision, the model can run at practical speeds on a single consumer-grade GPU in a Mac or PC, and it leads open-source models of the same scale in tool use and multi-step planning. The report covers the model architecture, distillation recipe, quantization, and more.

    #Papers/Research#Open-source ecosystem#English text is being prepared.

  6. Ma Dongxi NLPResearch
    AI score75

    Meta releases technical report for Muse Glimmer 30B, an agentic model for consumer-grade hardware

    Meta has released the technical report for Muse Glimmer, a 30B open-weight model aimed at agentic tasks on consumer-grade hardware; it runs at 4-bit precision on a single consumer-grade GPU and leads open models of the same scale on tool calling and multi-step planning. The relayer adds that the model surpasses Qwen3.6-27B and Gemma4-31B on some metrics, is about 17GB after 4-bit quantization, targets devices with 24GB of VRAM, and uses DFlash speculative decoding to reach about 233 tokens/s on an RTX 5090, roughly 3.1 times faster than ordinary decoding.

    Quoted postAI at Meta · @AIatMeta

    Today we published a technical report on Muse Glimmer, our 30B open-weight model built for agentic tasks on consumer hardware. It runs at practical speeds on a Mac or PC with a single consumer GPU at 4-bit precision, and leads comparably-sized open models on tool use and multi-step planning. Read about the model architecture, distillation recipe, quantization and more: https://t.co/H7n9qeDhxn

    #Model releases#English text is being prepared.#Agents

  8. Rohan PaulResearch
    AI score56

    Princeton and Google Paper: 4B Decision Model Shares Agent Tool Selection, Cutting Costs by 81.1% to 89.0%

    A new paper from Princeton University and Google finds that the most accurate decision model on isolated problems solved only 69.5% of agent tasks, while a 4B model trained on agent states solved 91.5%. Its harness delegates each step's tool selection to a fast decision model, adopting its answer only when confidence exceeds a threshold fitted on the agent's own runs, and it trained AgentDecision-4B on the states the agent actually reached (including states after mistakes). On tasks similar to the training data, such 4B models cut agent costs by 81.1% to 89.0% and boost speed by up to 2.5x, with a success-rate gap within 3 percentage points; without a threshold, delegating all tool selection drops the success rate by 19.0 percentage points. The paper is available at arxiv.org/abs/2610.21412.

    #Papers/Research#Agents#Reasoning

  9. Rohan PaulResearch
    AI score62

    Salesforce paper finds that cross-party agents do not proactively verify each other's claims

    A new Salesforce paper finds that agents representing different stakeholders do not verify their counterparts' claims when not prompted, so teams inherit the errors of the weakest agent. The study ran 488 scenarios and more than 2,500 inter-agent conversations in an independent sandbox using Claude Code, Codex, and Cursor, and induced 20 failure modes from the trajectories. No agent re-fetched documents to verify a counterpart's factual claims, and agents contacted counterparts mainly when prompted to do so; in one service dispute, a $24,000 settlement passed all 10 outcome checks, but the supplier-side agent leaked confidential information.

    #Papers/Research#Agents#Evaluations / benchmarks