English translation is being prepared.
Matched fieldsSummary
←Search results
Search “Reasoning”
English translation is being prepared.
Matched fieldsSummary
OpenAI explains how to detect behavioral misalignment in frontier models
OpenAI has disclosed its frontier model misalignment monitoring system, which covers training, evaluation, internal use, and some external deployments, and employs a four-layer architecture consisting of activation probes, model review, investigation agents, and automated investigation. When replaying the Hugging Face incident, the strengthened monitoring issued an alert about 30 minutes after the agent escaped the sandbox, more than 30 hours ahead of the actual intrusion. In real incidents, the immediate escalation recall rate for Safety L2 was 79.6%, and the recall rate for the Alignment L2 queue was 91.4%.
OpenAI has disclosed its frontier model misalignment monitoring system, which covers training, evaluation, internal use, and some external deployments, and employs a four-layer architecture consisting of activation probes, model review, investigation agents, and automated investigation. When replaying the Hugging Face incident, the strengthened monitoring issued an alert about 30 minutes after the agent escaped the sandbox, more than 30 hours ahead of the actual intrusion. In real incidents, the
Matched fieldsBody
Google Research releases SLEEP: using offline sleep cycles to mitigate catastrophic forgetting in large models
Google Research releases the SLEEP method, which lets LLMs periodically switch between wakeful interaction and sleep consolidation, writing short-term contextual information into scalable long-term parameters, and self-improving with synthetic data during the dreaming phase. The sleep phase unlocks reserved MoE memory blocks, combining generalized knowledge distillation and imitation learning to complete memory consolidation, then resets high-frequency temporary parameters. In evaluations of continual translation on unseen languages, SLEEP shows near-zero performance drop; on BABILong it is claimed to scale to about 10 million tokens; on AIME 24/25 and HMMT 25 it achieves approximately 79.2%, 69.0%, and 46.1%, respectively.
Google Research releases the SLEEP method, which lets LLMs periodically switch between wakeful interaction and sleep consolidation, writing short-term contextual information into scalable long-term parameters, and self-improving with synthetic data during the dreaming phase. The sleep phase unlocks reserved MoE memory blocks, combining generalized knowledge distillation and imitation learning to complete memory consolidation, then resets high-frequency temporary parameters. In evaluations of con
Matched fieldsBody
Arena analyzes 26,225 Agent sessions: without caching, agent costs are 3.6 times higher
Arena analyzed 26,225 Agent sessions, 30 agents, and 13 labs from September 12 to October 23, 2026, and found that if billed at the full input price, the typical task cost for the top ten agents would be 3.6 times higher, and 592% higher on MiMo V2.6 Pro. For agents, 86% to 96% of tokens are repeatedly read context, while their own output accounts for only 1% to 2%. Cache invalidation increases the typical task cost by about 5%, and on Claude Fable 5.1, 1 in every 10 sessions sees an increase of more than 18%.
Arena analyzed 26,225 Agent sessions, 30 agents, and 13 labs from September 12 to October 23, 2026, and found that if billed at the full input price, the typical task cost for the top ten agents would be 3.6 times higher, and 592% higher on MiMo V2.6 Pro. For agents, 86% to 96% of tokens are repeatedly read context, while their own output accounts for only 1% to 2%. Cache invalidation increases the typical task cost by about 5%, and on Claude Fable 5.1, 1 in every 10 sessions sees an increase of
Matched fieldsBody
AgiBot, together with Tsinghua University and Infinigence AI, releases the CEL collective experience learning paradigm
AgiBot, together with Tsinghua University and Infinigence AI, has released the collective experience learning paradigm CEL, which enables robots to use their peers' real operational trajectories as learning material, with humans only responsible for judging outcomes. On five real-machine tasks—flower arranging, clothes folding, phone charging, stool assembly, and phone packaging—the policies trained with CEL all exceeded the average level of 6 human teleoperators, leading by up to 16.7 percentage points, with an overall success rate of 97.5%; in the phone packaging task, the number of robots participating in training expanded from 16 to 96, and the success rate rose from 86.7% to 98.3%. The accompanying FERT algorithm uses failure trajectories to identify weaknesses and retrieves successful segments for prioritized sampling, while the FATS fault-tolerant asynchronous training system completed a cumulative 9,834 rounds of task collection in the 96-robot experiment, with the machine-time loss caused by 30 failures accounting for only 1.3% of the cluster's total operating time.
AgiBot, together with Tsinghua University and Infinigence AI, has released the collective experience learning paradigm CEL, which enables robots to use their peers' real operational trajectories as learning material, with humans only responsible for judging outcomes. On five real-machine tasks—flower arranging, clothes folding, phone charging, stool assembly, and phone packaging—the policies trained with CEL all exceeded the average level of 6 human teleoperators, leading by up to 16.7 percentag
Matched fieldsBody
Google research team proposes Sleep Framework, enabling large models to learn while in use without forgetting old knowledge
The Google research team has released a continual learning study called the "Sleep Framework," which divides model operation into two phases—awake and sleep—to address the problem of large models' rigid knowledge and difficulty in absorbing new information. The sleep phase writes short-term context into long-term parameters through parameter expansion, mixture-of-experts memory blocks, and knowledge distillation, and generates synthetic samples via "dreaming" for reinforcement learning. Experiments show that the framework can handle contexts of up to about 10 million tokens, that performance degradation is close to zero in multi-stage continual learning on unseen languages, and that it achieves higher accuracy on mathematical benchmarks such as AIME 24/25.
The Google research team has released a continual learning study called the "Sleep Framework," which divides model operation into two phases—awake and sleep—to address the problem of large models' rigid knowledge and difficulty in absorbing new information. The sleep phase writes short-term context into long-term parameters through parameter expansion, mixture-of-experts memory blocks, and knowledge distillation, and generates synthetic samples via "dreaming" for reinforcement learning. Experime
Matched fieldsBody
Fangcun Lab releases an 87-page safety evaluation report on DeepSeek V4.1 Flash-1017
The third-party evaluation lab Fangcun (@FangcunLeap) from China has released an 87-page report evaluating DeepSeek V4.1 Flash-1017's performance on loss of control and harmful manipulation. The report finds that it shows clear safety progress compared with the DeepSeek V4 generation, approaching or even matching the top closed-source models tested in the report on points such as "not strategically hiding capabilities and still respecting approval boundaries under work pressure." Shortcomings include overly strong evaluation awareness, more frequent overreach in long multi-step tasks than the previous V4.1 Flash generation, greater stance drift under multi-turn persuasion than the GPT in the report, and executing in most conversations when the system prompt asks it to mislead users.
The third-party evaluation lab Fangcun (@FangcunLeap) from China has released an 87-page report evaluating DeepSeek V4.1 Flash-1017's performance on loss of control and harmful manipulation. The report finds that it shows clear safety progress compared with the DeepSeek V4 generation, approaching or even matching the top closed-source models tested in the report on points such as "not strategically hiding capabilities and still respecting approval boundaries under work pressure." Shortcomings in
Matched fieldsBody
Anthropic study: experts adopt AI agents' random preferences and mistake them for their own judgments
Anthropic ran an experiment with 50 causal inference experts; the agent's preference at each decision point was randomly assigned by a coin flip, yet participants still adopted the agent's preference in 86% of analytical decisions, with flip rates of 62% to 82%. Afterward, 52% believed the final recommendation was more like their own judgment, 35% believed it was basically or entirely their own, and only 15% mentioned AI. The study also found that after using the agent, the rate of serious analytical errors dropped from 55% (6/11) to 6% (3/50), and the median time required fell from 90 minutes to 65 minutes, but the no-AI group had a smaller sample and slightly lower expertise.
Anthropic ran an experiment with 50 causal inference experts; the agent's preference at each decision point was randomly assigned by a coin flip, yet participants still adopted the agent's preference in 86% of analytical decisions, with flip rates of 62% to 82%. Afterward, 52% believed the final recommendation was more like their own judgment, 35% believed it was basically or entirely their own, and only 15% mentioned AI. The study also found that after using the agent, the rate of serious analy
Matched fieldsBody
Meta Puts JiTTests into Production: LLM-Generated Tests Catch Defects Before Code Is Merged
Meta has published the production deployment results of Catching JiTTests, a system that uses LLMs to automatically generate executable tests for pending code changes, exposing defects before the code is merged. The system employs two workflows, intent-aware and dodgy diff, and uses an LLM-as-Judge ensemble together with a rule-based method, RubFake, to filter false positives, while reducing manual line-by-line review through a human-on-the-loop approach. Meta says 589 engineers have participated, the true positive rate is 60%, and in the first half of 2026 it caught 74 severe privacy SEVs at the diff stage, reducing the overall incidence of such issues by 20%.
Meta has published the production deployment results of Catching JiTTests, a system that uses LLMs to automatically generate executable tests for pending code changes, exposing defects before the code is merged. The system employs two workflows, intent-aware and dodgy diff, and uses an LLM-as-Judge ensemble together with a rule-based method, RubFake, to filter false positives, while reducing manual line-by-line review through a human-on-the-loop approach. Meta says 589 engineers have participate
Matched fieldsBody
Google Research proposes the sleep framework, enabling LLMs to retain knowledge during continual learning
Google Research proposes the sleep framework, which introduces periodic sleep stages to LLMs to consolidate short-term context into long-term parameters, thereby mitigating catastrophic forgetting and static knowledge cutoff. The framework consists of two stages: memory consolidation and dreaming. The former uses MoE parameter expansion and knowledge seeding based on GKD and reinforcement learning imitation, while the latter achieves self-improvement through random expert activation, filtering synthetic samples, and RL refinement. On BABILong, this method scales nearly perfectly within 10M tokens, and achieves accuracies of 79.2%, 69.0%, and 46.1% on AIME 24/25 and HMMT 25, respectively.
Google Research proposes the sleep framework, which introduces periodic sleep stages to LLMs to consolidate short-term context into long-term parameters, thereby mitigating catastrophic forgetting and static knowledge cutoff. The framework consists of two stages: memory consolidation and dreaming. The former uses MoE parameter expansion and knowledge seeding based on GKD and reinforcement learning imitation, while the latter achieves self-improvement through random expert activation, filtering s
Matched fieldsBody
Cartesia Research: From Tokens to Bytes, How Tokenization Priors Affect Language Model Downstream Transfer
Cartesia has released the second article in its series, using two types of scaling laws, compute-to-loss and loss-to-loss, to compare byte-level Transformers, BPE, BPE-NP with pre-tokenization removed, BLT-like, and H-Net. The conclusion is that BPE tokenization is nearly 5 times more compute-efficient than byte-level models in pretraining, but at the same validation bpb, OLMES downstream loss is significantly worse, equivalent to a compute gap of about 1.9 times; H-Net simultaneously obtains the compute efficiency brought by compression and the transfer capability of byte-level models, with a direct compute-to-OLMES fit of about 2.06 times. The article also points out that the manual regularization step of pre-tokenization has a significant impact on both metrics, and discusses the prompt boundary issue of BPE.
Cartesia has released the second article in its series, using two types of scaling laws, compute-to-loss and loss-to-loss, to compare byte-level Transformers, BPE, BPE-NP with pre-tokenization removed, BLT-like, and H-Net. The conclusion is that BPE tokenization is nearly 5 times more compute-efficient than byte-level models in pretraining, but at the same validation bpb, OLMES downstream loss is significantly worse, equivalent to a compute gap of about 1.9 times; H-Net simultaneously obtains th
Matched fieldsBody
UK AISI Releases ReproBAIT Evaluation: Can AI Coding Agents Reproduce Biological AI Tools from Papers?
The UK AI Security Institute's Chem-Bio team, in collaboration with SecureBio, built the ReproBAIT evaluation to test the ability of AI coding agents to reproduce biological AI tools (BAITs) given data and limited public information. Results showed that simpler tools could be reproduced within 4 hours and under a $20 compute budget; ProteinMPNN reached 96.5% reproduction fidelity with Claude Opus 5 when the full paper was provided, while the most complex tool, RFdiffusion3, was not successfully reproduced by any agent within the time limit. Based on this, the team noted that merely withholding code or restricting paper information does not equal safe release, and that controlling access to underlying biological data may be more effective than controlling code.
The UK AI Security Institute's Chem-Bio team, in collaboration with SecureBio, built the ReproBAIT evaluation to test the ability of AI coding agents to reproduce biological AI tools (BAITs) given data and limited public information. Results showed that simpler tools could be reproduced within 4 hours and under a $20 compute budget; ProteinMPNN reached 96.5% reproduction fidelity with Claude Opus 5 when the full paper was provided, while the most complex tool, RFdiffusion3, was not successfully
Matched fieldsBody
ModelBest and Tsinghua Release PostTrainBench: Agents Execute Well but Are Strategy-Locked
The ModelBest and Tsinghua team analyzed 1,338 complete training trajectories based on PostTrainBench, covering 7 benchmarks, 4 base models, and Claude Code, Codex CLI, and OpenCode, running for 10 hours on a single H100. Results show that the agents' average score rose from 33.34 to 48.55 with experience, but among 3,557 pairs of adjacent experiments, only 2.1% actually switched strategies; all execution-layer suggestions were adopted, while none of the strategy-layer suggestions were adopted. At the same checkpoint, a single sentence of human guidance to switch strategies was enough to outperform autonomous continuation, with GSM8K improving by 17.44pp, HumanEval by 10.80pp, and AIME 2025 by 6.67pp.
The ModelBest and Tsinghua team analyzed 1,338 complete training trajectories based on PostTrainBench, covering 7 benchmarks, 4 base models, and Claude Code, Codex CLI, and OpenCode, running for 10 hours on a single H100. Results show that the agents' average score rose from 33.34 to 48.55 with experience, but among 3,557 pairs of adjacent experiments, only 2.1% actually switched strategies; all execution-layer suggestions were adopted, while none of the strategy-layer suggestions were adopted.
Matched fieldsBody
Convovo study: When forced to choose between two options, Grok 4.7 picks Republicans nine times out of ten, while GPT-6.1 Sol picks Democrats 97% of the time
A Convovo study of nearly 500 elections found that when forced to choose between two options via API, SpaceXAI's Grok 4.7 recommended the Republican candidate in 92% of responses, while OpenAI's GPT-6.1 Sol recommended the Democratic candidate in 97% of responses, with the two agreeing less than 10% of the time. The study also found that Grok frequently cited conservative sources, and that each API request carried about 1,240 additional tokens that researchers could neither view nor remove; after web search was disabled, GPT's Democratic lean persisted, while Grok's Republican lean disappeared, with 27 of 30 cases declining to choose. OpenAI responded that the test targeted the developer API and that the prompt explicitly required recommending a candidate, so it does not reflect the ChatGPT consumer experience, and that the consumer app would refuse to answer such questions.
A Convovo study of nearly 500 elections found that when forced to choose between two options via API, SpaceXAI's Grok 4.7 recommended the Republican candidate in 92% of responses, while OpenAI's GPT-6.1 Sol recommended the Democratic candidate in 97% of responses, with the two agreeing less than 10% of the time. The study also found that Grok frequently cited conservative sources, and that each API request carried about 1,240 additional tokens that researchers could neither view nor remove; afte
Matched fieldsBody
Harvard study says AI review comments comprehensively outperform high-level human peer reviewers
Two researchers from Harvard University published a paper on October 22, comparing 134 pairs of AI and human review comments; AI led comprehensively across seven evaluation dimensions. Three experts who had served as editors-in-chief of major academic journals conducted 30 blind evaluations, and in 26 of them they judged the AI review comments to be better. The study used 124 papers and 134 paid professional review comments provided by rapidPeer, with Claude Opus 4.8 generating the reviews and GPT-5.5 serving as the blind-evaluation judge. The factual error rate of AI review comments was 5.5%, compared with 15.9% for humans.
Two researchers from Harvard University published a paper on October 22, comparing 134 pairs of AI and human review comments; AI led comprehensively across seven evaluation dimensions. Three experts who had served as editors-in-chief of major academic journals conducted 30 blind evaluations, and in 26 of them they judged the AI review comments to be better. The study used 124 papers and 134 paid professional review comments provided by rapidPeer, with Claude Opus 4.8 generating the reviews and G
Matched fieldsBody
Harvard study: AI review comments comprehensively outperform high-level human peer reviewers across seven dimensions
Two researchers from Harvard University published a paper on October 22 comparing 134 pairs of AI and human review comments, with AI leading comprehensively across seven evaluation dimensions. The study used Claude Opus 4.8 to generate review comments and GPT-5.5 as the blind-review judge; in all 134 comparisons GPT-5.5 chose the AI. In 30 blind reviews by three former academic journal editors-in-chief, 26 considered the AI better. The factual error rate of AI review comments was 5.5%, versus 15.9% for humans, and all 651 references cited by the AI with DOI or arXiv numbers actually exist.
Two researchers from Harvard University published a paper on October 22 comparing 134 pairs of AI and human review comments, with AI leading comprehensively across seven evaluation dimensions. The study used Claude Opus 4.8 to generate review comments and GPT-5.5 as the blind-review judge; in all 134 comparisons GPT-5.5 chose the AI. In 30 blind reviews by three former academic journal editors-in-chief, 26 considered the AI better. The factual error rate of AI review comments was 5.5%, versus 15
Matched fieldsBody
CAICT report: Large model token prices doubled compared with the second half of last year, average daily consumption up 3x
The Institute of Artificial Intelligence of the China Academy of Information and Communications Technology and other institutions jointly released the "Large Model Token Service Performance Monitoring Report (2026)", showing that domestic token prices doubled compared with the second half of last year, with the normalized average price rising from 6.8 yuan to 13.46 yuan, while in June 2026 domestic average daily token consumption was about 500 trillion, nearly 4 times that of December 2025. In terms of performance, in August the average TPS of domestic models reached 78.23, doubling the 2025 average of 38.25, but the average TPS of typical foreign models was 112.82; the domestic first-token latency rose to 1.26 seconds. The report also compared four open-source models, DeepSeek, Qwen, GLM and Kimi, and found that the TPS gap for the same model across different platforms can be close to a factor of two.
The Institute of Artificial Intelligence of the China Academy of Information and Communications Technology and other institutions jointly released the "Large Model Token Service Performance Monitoring Report (2026)", showing that domestic token prices doubled compared with the second half of last year, with the normalized average price rising from 6.8 yuan to 13.46 yuan, while in June 2026 domestic average daily token consumption was about 500 trillion, nearly 4 times that of December 2025. In t
Matched fieldsBody
W3C discussion draft proposes using Web standards to constrain AI agents
The W3C Strategy team published a discussion draft, "W3C and the State of the Web in the AI Era," proposing the use of Web standards to make AI agents safer, focusing on Prompt Injection protection, controllable permissions, and accessibility. The article was published on October 21 to build momentum for the TPAC meeting; it is a proposal rather than an approved standard, with the official presentation scheduled for October 29. Using travel booking as an example, the article points out that letting an agent search for flights does not equal authorizing it to pay for tickets, and recommends setting spending limits, revocable authorization, and traceable credentials for payments.
The W3C Strategy team published a discussion draft, "W3C and the State of the Web in the AI Era," proposing the use of Web standards to make AI agents safer, focusing on Prompt Injection protection, controllable permissions, and accessibility. The article was published on October 21 to build momentum for the TPAC meeting; it is a proposal rather than an approved standard, with the official presentation scheduled for October 29. Using travel booking as an example, the article points out that lett
Matched fieldsBody