AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38372v1 Announce Type: new Abstract: A harness is the code around a language-model agent that organizes prompts, calls tools, manages context, and controls execution. As models grow stronger, recent work has begun to let agents improve their own harnesses, a line of work known as self-evolving harnesses. In most existing methods, a separate proposer running on a human-designed harness modifies the solver's harness, and a separate harness is evolved for each benchmark. Real-world tasks come from many domains, so both the evolution and the evaluation of a harness should cover a diverse range of tasks. We propose a framework close to recursive self-improvement: the same frozen model, on the same version of the harness, first solves tasks as the solver a…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38369v1 Announce Type: new Abstract: We study generalized Blaschke curves as a controlled environment for AI-assisted mathematical rediscovery. For one fixed degree-four Blaschke product, an agent receives numerical coordinates of the six pair-lines determined by each of 80 boundary configurations. The target theorem is withheld from the task instructions. The saved research log reports rejected geometric hypotheses and a homogeneous cubic fitted to polygon sides. Its frozen coefficients predict 480 lines from 80 unseen parameter values, with a recorded RMS scale-free residual of $8.88\times10^{-17}$. Discovery-set diagonals provide an out-of-fit consistency check, not a fully held-out test. A separate one-configuration run reports insufficient evide…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38359v1 Announce Type: new Abstract: High quality synthetic data is central to post training LLMs for adaptive AI applications that represent the diverse expert strategies and decisions in conversations. Prompting LLMs directly or conditioning them on end use scenarios yields low diversity data that collapses onto dominant modes. We propose a method to generate diverse high quality synthetic data using Generative Flow Networks (GFlowNets). We show that training GFlowNets to generate latent conversation structure using a Gaussian mixture density over key interaction features (e.g., confusion episode dynamics, scaffolding directive balance) enables sampling expert strategies in proportion to their prevalence in the training data. Across two structurall…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38346v1 Announce Type: new Abstract: When a student is stuck, a tutor faces the assistance dilemma: help given too early can hinder productive struggle, while help withheld too long leaves the student in a frustrating, persistent impasse (i.e., wheel spinning). Generative AI tutors increasingly use guardrails restricting answer-giving, yet little is known about how such tutors behave once an impasse persists. We analyze 20,462 student turns from 1,260 authentic sessions with a guided LLM chemistry tutor, identifying 6,630 impasse turns of three major types: conceptual errors, expressed uncertainty, or help-seeking. We then used these impasses to simulate three tutoring conditions to study variation in AI tutor guidance through impasses: baseline, no-…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38340v1 Announce Type: new Abstract: When a materials LLM answers a question about crystal structure, does it reason from the structure or copy an answer already printed in its input? Accuracy cannot tell: a structural description often prints the very field it is scored against. CARAT holds question and gold answer fixed across eight matched views, names each structural relation separately in GraphSpace, and adds matched fine-tuning, answer masking, evidence injection, paired inference, and a rule that can withhold claims. First, on the benchmark's hardest families the grounded view is worth 17.3 points over formula inputs. Second, we turn that scrutiny on ourselves. GraphSpace beats a plain periodic graph by 19.3 points, but that margin is two effe…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38296v1 Announce Type: new Abstract: Large language models (LLMs) can influence people's beliefs, yet little is known about whether and how they can manipulate each other. To investigate this, we simulate conversations between two agents: a target LLM that role-plays a human persona based on demographic and psychological attributes, and an influencer LLM that aims to make the target's beliefs more extreme. We examine radicalization along two pathways: resonance, where the influencer reinforces a target's pre-existing belief, and persuasion, where the influencer promotes a belief the target initially considers unimportant. Across affective and behavioral metrics, we find that both mechanisms radicalize the target. However, resonance produces consisten…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38294v1 Announce Type: new Abstract: We study the generation of agentic workflows that jointly optimize multiple objectives, such as accuracy, cost, latency, robustness, and consistency. Existing methods for workflow generation typically optimize accuracy alone or a weighted sum of objectives, so each trained generator commits to one fixed trade-off and must be retrained from scratch when preferences change. To alleviate this, we propose MoFlow, which generates workflows optimized across varied preferences. Specifically, MoFlow formulates workflow generation as a multi-objective Markov decision process and solves it by leveraging Convex-Hull Monte Carlo Tree Search with optimistic set-valued backups, where every node stores a set of reachable trade-o…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38288v1 Announce Type: new Abstract: We present AREX-2, an effort to advance the self-improving capability of LLM agents, which we define as the ability to iteratively refine a solution at test time. This ability rests on two complementary capabilities: reflection, which produces a solution better than the current one, and long-horizon execution, which keeps the iteration effective over many rounds. We hypothesize that both capabilities are domain-agnostic, and can therefore be learned in scenarios that are well suited for supervision. Accordingly, we synthesize long-horizon improvement trajectories from machine learning and algorithmic programming tasks, two domains that offer verifiable feedback and reward sustained iteration. Trained on this data,…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：arXiv:2609.38282v1 Announce Type: new Abstract: Vision-language models may rewrite anomalous text in images into linguistically plausible expressions, compromising OCR transcription faithfulness. Sequence-level task rewards and local teacher guidance are complementary, but guidance from the same teacher may not remain equally effective as the student improves. Offline analysis shows that supervision from a fixed teacher becomes progressively less favorable as the student improves, both across training checkpoints and across response groups with different task rewards. Motivated by this observation, we introduce GAD-RL, which adaptively regulates teacher supervision during joint post-training according to the student's current task performance and local distribu…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Perplexity Research and turbopuffer have released pplx-embed-v2-context-9b-preview, a contextual embedding model for RAG pipelines. Each chunk is embedded with the full document in view. The real change is the training signal. The model learns to retrieve the answer along with the context needed to verify it, not one ‘gold passage.’ P Is it deployable? Yes, […] The post Perplexity Releases pplx-embed-v2-context-9b-preview: A Contextual Embedding Model That Retrieves Answers and Their Supporting Evidence appeared first on MarkTechPost.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Some experts fear AI could be weaponised in south-east Asia, home to nations with predominately young and hyper-connected populations Years ago it would have been an elaborate operation involving any army of cyber troops creating fake news websites, social media accounts and forged dossiers, all deployed to spread disinformation en masse. Now, all you need is AI. Continue reading...

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：IP data is warehousing dataEvery firewall, load balancer, VPN, CDN edge, DNS resolver,...

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Grokipedia, SpaceXAI's AI-powered competitor to Wikipedia, recently started incorporating edits again, and today, it got some design tweaks as part of a v0.3 update, including a new logo and refreshes to its homepage and live edits page. SpaceXAI head of design Benji Taylor calls it a "newly refreshed Grokipedia." Grokipedia's old homepage was pretty much just a logo and a search bar, but the updated homepage includes a list of featured articles, a list of the most-read articles, and a tracker showing the "latest edits." The new featured list showcases digital representations of books that you can spin around, and the most read section also … Read the full story at The Verge.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Democratic governor sharply critical of Donald Trump for not passing comprehensive federal AI regulations Governor Gavin Newsom signed laws Wednesday aimed at protecting California workers from the threats of artificial intelligence , including potential job losses and workplace surveillance. The laws ban employers from using the technology to predict a worker’s emotional state by using their biometric data, require employers to send written notices to workers if AI is responsible for mass layoffs and ban employers from relying on AI to decide to fire someone. Continue reading...

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Sidecars are isolated containers that run alongside your main Sandbox on the same host.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Everything announced live at Modal's inaugural conference, Runtime.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：As it opens a new location, the social club prepares grant applications in 2 hours instead of 3 days and liquor-license materials in 3 hours instead of 4 days.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：VM Sandboxes are built for those who need to give their agents the power of a full computer.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Weaviate v1.39.3 fixes a high-severity credential disclosure issue in the Google modules.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：The common paradigm of reinforcement learning with verifiable rewards (RLVR) is to let agents make multiple attempts at a task, and optimize towards the successful ones. This becomes problematic in the realms of self-improvement, where tasks are so difficult that the agent has a low or even no chance of success, and where there are no teacher models or example solutions to distill from. In this paper, we introduce RLTL;DR. After each failed attempt, we show the policy the verifier outputs and let it write its own feedback, in the form of a single TL;DR insight. The next rollout is conditioned…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Recent autonomous machine learning engineering (MLE) agents have made significant progress on public leaderboards. Often motivated by progress stagnation over long-horizon cycles and limited Large Language Model (LLM) primitives, modern MLE agents are deployed on top of increasingly elaborate machinery: multi-agent orchestrators, dedicated retrieval subagents, and more. While such harnesses expand, the use of more primitive but improved coding agents—where LLMs have direct access to the execution environment through read, write, and bash primitives—has received little attention in the field…

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Discussion | Link

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Our DevDay coverage - the first pod on the DevDay lineup - dives in with the leaders of OpenAI’s CUA team and API platform.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Discussion | Link

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯： I visited the Museum of the City of New York today and got to see He Built This City: Joe Macken’s Model, the 50 x27 feet model of the city built over a 21 year period from balsa wood and cardboard. It exceeded my already high expectations. The exhibition closes on 12th October so you should absolutely make a priority to see it if you get the chance. Tags: museums, new-york

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Gemini 4 Argon tops GPT-6 Astra and Claude Opus 5.5 on most benchmarks, but access remains gated today. The post Google DeepMind Unveils Gemini 4 Argon with 1M Output Tokens for Coding, Knowledge Work and Cyber Defense appeared first on MarkTechPost.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Discussion | Link

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Discussion | Link

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：OpenAI released GPT-6.1 Sol on September 29, 2026, an upgrade to GPT-6 Sol. It reaches near-Astra results on agentic coding, computer use and professional work at one-fifth of Astra's token prices. It costs $2 input and $10 output per million tokens, and cached input drops to $0.10. It is available now through the OpenAI API, ChatGPT Work and Codex. The post OpenAI Releases GPT-6.1 Sol: Near-Astra Coding and Computer Use at One-Fifth of Astra’s Token Price appeared first on MarkTechPost.

AI 服務暫時不可用，以下為來源摘要，待恢復後補全翻譯：Google today revealed its next AI frontier model, which it's calling Gemini 4 Argon. The new model delivers "frontier performance in complex workflows across real-world software engineering, enterprise knowledge work like legal and finance, and cybersecurity defense," according to chief AI architect and Google DeepMind SVP Koray Kavukcuoglu. But the company is limiting access at first to a "set of trusted cyber defenders," and Kavukcuoglu says that Google is "actively engaged in the U.S. government's voluntary process for pre-release model access while we gradually expand access." Kavukcuoglu says Gemini 4 Argon is already powering Google's … Read the full story at The Verge.