New Model Releases A

An arXiv paper presents a systematic analysis of how LLMs can learn, via reinforcement learning, to adapt their query formulation strategies to different retrievers in retrieval-augmented generation. Summarized neutrally from the abstract.

The paper introduces GIST-CMTF, which augments Causal Minimal Tool Filtering with goal-state inference for tool-augmented LLM agents. It addresses wrong-goal execution, where ambiguous requests such as "handle my appointment" map to multiple goals and an agent may follow a valid causal tool path toward an unintended objective.

The paper presents MIXGUARD, a mixup-based privacy-preserving split-learning framework for LLMs combining token- and representation-level obfuscation with adaptive gradient perturbation to balance utility, privacy, and efficiency. Claims reflect the abstract.

The paper introduces MST-CLIPIQA, a multi-scale two-stream framework for assessing AI-generated image quality. It argues that monolithic vision-language representations entangle semantic understanding with low-level perceptual sensitivity, and instead decouples them using dual CLIP encoders for hierarchical alignment.

The paper proposes Collective Skill Tree Search (CSTS), a tree-search framework that automatically builds reusable skills for LLM agents via iterative collective generation and assessment across multiple models. Claims reflect the abstract.

As LLM post-training RL uses multi-dimensional rewards, conflicting signals across reward groups can cancel out and hinder training. GD²PO decouples rewards into groups and, inspired by DAPO, dynamically filters near-zero-advantage rollouts, reducing conflicts and improving RL training efficiency.

The paper introduces P3B3, an expert-curated benchmark and framework for measuring European versus Brazilian Portuguese variety bias in LLMs. It reports most models lean strongly toward pt-BR and argues for more balanced multilingual representation.

MyPCBench evaluates computer-use agents as personal assistants on a Linux desktop with 17 simulated web apps and 184 persona-seeded tasks, benchmarking six closed and open-weight models. Reported scores reflect the paper and are not independently verified.

The paper proposes detecting AI-generated videos, especially those from text-to-video models, by amplifying noise to reveal subtle artifacts that distinguish them from authentic footage. It notes that prior work largely targeted GAN-generated samples and frames text-to-video detection as still underexplored.

The paper injects intent-based misinformation into single- and multi-agent LLM systems and finds it degrades performance and persists through debate, though multi-agent debate can reduce degradation when most agents are uncontaminated. Robustness depends on group composition.

The paper introduces Reflective Masking, a lightweight post-training method that lets mask diffusion models iteratively revisit and revise prior outputs via multi-turn masking, plus a History Reference component. Claims reflect the abstract and are not independently verified.

Note: the abstract was unavailable, so this is summarized neutrally from the title alone. The paper proposes SCAR, a 'semantic continuity-aware retrieval' method aimed at efficient context expansion in retrieval-augmented generation (RAG). Specific mechanisms and evaluation results cannot be confirmed from the title.

The paper introduces FraudSMSWalker, a controlled benchmark for URL-masked SMS-to-webpage fraud judgment. It contains 699 bilingual chains (332 fraudulent, 367 benign) across ten scenarios, withholding raw URLs, hosts, and reputation metadata so models cannot rely on reputation shortcuts, and evaluates nine web agents.

This arXiv paper introduces VeriGraph, a traceable neuro-symbolic reasoning framework for verifiable data-analytic agents. The authors note that LLM agents' reliance on linear text trajectories makes reasoning hard to audit, entangling deterministic computations over raw data with semantic deductions over natural-language claims. VeriGraph instead has agents build an explicit heterogeneous evidence directed acyclic graph (DAG) during execution.

This arXiv paper introduces two complementary benchmark datasets for computational lexical semantic change (LSC) detection. The Bi-Directional Lexical Semantic Change (BD-LSC) dataset captures sense gain, loss, and stability across three time periods, targeting cases—especially slang versus standard usage—where words simultaneously gain and lose senses, which existing benchmarks struggle to capture.

On its 40th anniversary, the Japanese Society for AI issued proposals for adopting AI across Japanese society. Asserting that AI will not replace humans, it offered four recommendations and touched on issues spanning security and copyright.

IBM's new AI tool 'IBM Bob' reportedly cut Java application modernization from 30 days to 3 at early adopters. Its distinguishing feature is going beyond mere source-code generation.

Sakana AI has launched Sakana Marlin, an AI research agent, commercializing the beta it had offered since April. Ahead of the release it held a press hands-on, showing reporters reports the AI generated from pre-collected themes.

Researchers at Georgia Tech, the University of Michigan and others published a study comparing whether AI chatbots or search engines yield better learning. Over an eight-day experiment, the paper examines how generative AI shapes information seeking and learning.

OpenAI introduced its Partner Network, committing $150M to help global partners accelerate enterprise AI adoption, deployment, and transformation. The program aims to broaden OpenAI's reach into enterprise markets through a structured partner ecosystem.

Research firm Gartner says the enterprise AI coding-agent market has entered a new phase of growth and competitive realignment. It predicts that by 2027, 65% of teams coding with AI agents will no longer regard an IDE as essential.

Sakana AI has launched Sakana Marlin, its first commercial product: an autonomous research assistant for business. Given a research theme, it works autonomously for up to about eight hours—forming hypotheses, gathering and verifying information—then outputs structured summary slides and a report spanning dozens of pages. Built on the firm's long-horizon reasoning technology, it aims to act as a 'virtual CSO,' is self-serve, and available same day, with plans from free pay-per-use to Enterprise.

Amazon CEO Andy Jassy was among the tech executives who raised security concerns about Anthropic's frontier models to Trump administration officials, sources told Reuters, ahead of the directive barring foreign nationals from using Fable 5 and Mythos 5.

An early 0.1a0 release of luau-wasm packages Luau, Roblox's typed Lua dialect, for WebAssembly, built using the newly enabled approach for publishing WASM wheels to PyPI for Pyodide.

On June 12 Anthropic said it would suspend its flagship Claude Fable 5 and Mythos 5 for all users after a US export-control directive barred foreign nationals from access on security grounds. Calling it a misunderstanding, the firm aims to restore service soon; other models are unaffected.

Simon Willison updated his browser tool for OpenAI's WebRTC realtime audio API. It now supports the newer realtime voice model touting GPT-5-class reasoning, and lets users paste document text as context for spoken conversations about it.

MONOist editors pick the week's top stories from June 8-12, led by Kioxia overtaking Toyota as Japan's most valuable listed company, and revisit 2005's top-10 market-cap ranking to trace shifts in Japan's corporate landscape.

Anthropic announced a partnership with Tata Consultancy Services. TCS will deploy Claude to 50,000 employees across 56 countries, build Claude-powered products for finance, healthcare and the public sector, and join the Claude Partner Network.

ClinHallu is a benchmark for diagnosing where hallucinations originate in medical multimodal LLM reasoning, decomposing traces into visual recognition, knowledge recall, and reasoning integration. It provides 7,031 validated instances and uses stage-replacement interventions to localize error sources.

AdaSR moves beyond the read-then-think paradigm by letting reasoning models reason incrementally as input streams in. It uses a hierarchical relative policy optimization scheme to train streaming reasoning.