AI News Today, Filtered for What Matters

What happened

LangChain introduced ReviewBench, a benchmark for evaluating code review agents against real pull request feedback from trusted reviewers.

What happened

Anthropic's Frontier Red Team published a report on three real-world incidents encountered during cybersecurity evaluations, describing lessons for testing, safeguards, and responsible disclosure.

What happened

OpenAI reported that enabling retained reasoning and compaction in its Responses API harness tripled GPT-5.6 Sol's ARC-AGI-3 public-set score and reduced output tokens by 6x.

What happened

arXiv:2607.24759v1 Announce Type: new Abstract: Research projects, educational efforts, and adjacent knowledge work accumulate findings, decisions, and reasoning that future collaborators rarely recover. The parts most useful to that work, including dead ends and walked-back cla

What happened

OpenAI published a field report on scientists using coding agents to modernize scientific software in genomics and other data-rich fields, emphasizing validation, maintenance, and human review.

What happened

arXiv:2607.19349v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly deployed as always-on online services, making efficient LLM serving a critical systems challenge. Achieving low latency and high throughput under volatile demand requires deep understan

What happened

arXiv:2607.16196v1 Announce Type: new Abstract: Soft, sensorized companions offer a physically safe and emotionally intuitive interface for socially assistive technologies, yet their deformability and multichannel tactile sensing complicate the robust interpretation of human aff

What happened

OpenAI published GPT-Red, an automated red-teaming system trained through self-play to find prompt-injection failures and improve GPT-5.6 robustness during model training.

What happened

We present AgentLens, a production-assessed benchmark for interactive code agents. Most code-agent benchmarks reduce a run to a single bit -- did the task pass? -- but the people who actually use these agents experience the entire trajectory: how the agent follows instructions, u

What happened

A new analysis from OpenAI reveals issues in SWE-Bench Pro, a popular coding benchmark, raising concerns about reliability and accuracy in evaluating AI models.

What happened

We study when natural-language feedback produces improvement beyond the gains obtainable from repeated attempts alone. In multi-turn language agent setting, higher final accuracy can reflect useful feedback, but it can also arise from resampling, format correction, or additional

What happened

Personality prompting shapes how large language models communicate, yet whether these behavioral shifts affect objective task outcomes remains under-explored. Prior work shows that agents prompted with low agreeableness produce adversarial language, while those prompted with high

What happened

Large Language Models (LLMs) have revolutionized language modeling through autoregressive generation, enabling strong performance across a wide range of tasks. Recently, Diffusion Language Models (DLMs) have emerged as an alternative paradigm that generates text through iterative

What happened

arXiv:2606.13686v1 Announce Type: new Abstract: As autonomous web agents are increasingly deployed to perform real-world tasks, ensuring their safety has become a critical concern. In this work, we study web agent behavior under realistic deceptive interfaces in the e-commerce d

What happened

NVIDIA says Blackwell led the first Agentic AI infrastructure benchmark, highlighting performance for agentic AI workloads.

What happened

A new arXiv paper argues that long-term AI agent memory should be treated as an evolving data-management workload, not just a collection of records, embeddings, or graph edges.

What happened

A new arXiv paper introduces SPEAR, an agentic prompt optimizer that can run Python analysis, evaluate prompts, revise them, and roll back when metrics regress.

What happened

OpenAI says one of its models contributed to disproving a central conjecture in the 80-year-old unit distance problem, with external mathematicians validating the result.

Related on ToolWorthy