Claude Sonnet 4.6

2026-11-02

RESEARCH · 1 SOURCE · arXiv cs.AI

ArXiv paper releases ProClaim and ProClaim-eval for consensus scientific claim verification

The paper (arXiv:2610.40906v1) introduces ProClaim, an agent instrument that persists a structured evidence state outside a model's context, and ProClaim-eval, a 718-claim dataset with SUPPORT/REFUTE gold labels from five expert-curated sources across molecular biology, nuclear physics, and astronomy. The authors find that maintaining an external evidence state and using a trained stopping judge improves consensus claim verification compared with in-context accumulation and fixed-round stopping, and that a separately trained judge matches or exceeds performance of some off-the-shelf Claude judges on parts of the dataset.

6.0

2026-10-31

MODELS · 1 SOURCE · The Decoder

Anthropic rolls out watermark to four Claude models, but two older models can be bypassed by changing sampling settings

Anthropic is adding its imperceptible text watermark to Claude Mythos 5, Claude Opus 4.7, Claude Opus 4.6, and Claude Sonnet 4.6; the watermark is embedded at the model level and will appear across hosting platforms (including Amazon Bedrock). On Opus 4.6 and Sonnet 4.6 the watermark only activates when developers leave sampling parameters (temperature, top_p, top_k) at their defaults, so tweaking any one parameter can evade the mark; Anthropic offers a private-preview detection API and cites the EU AI Act deadline as the reason for the rollout.

8.0

2026-10-27

COMPANIES · 1 SOURCE · AWS Machine Learning

Exowatt uses SageMaker serverless customization to fine-tune Qwen3 models for edge and cloud, beating frontier models

Exowatt built a repeatable pipeline on Amazon SageMaker AI serverless model customization to produce task-specific fine-tuned models for edge alarm triage (Qwen3-1.7B) and cloud energy-market analysis (Qwen3-14B). The tuned models outperformed benchmark frontier models on accuracy for these tasks and reached up to 49× lower cost, while enabling a repeatable path from raw operational data to deployable artifacts in days rather than months.

6.0

2026-10-21

RESEARCH · 1 SOURCE · WIRED AI

CrowdStrike finds criminal hackers bypassing AI guardrails using Claude and Deepseek

CrowdStrike researchers analyzed infrastructure left exposed from May–August by four threat groups and found attackers using agentic AI harnesses—largely older Anthropic Claude variants (including Claude Opus 4.8, 4.5, Haiku 4.5, Sonnet 4.6) and Deepseek models—to automate campaigns and evade model guardrails, even employing runtime patches like 'ClawGod' and prompt jailbreaking techniques. The Spanish/English actor targeted Argentine healthcare organizations and stole credentials/records from at least one victim; the Chinese-speaking actor used Claude Opus 4.8 to compromise government and private targets in Brazil and Colombia among other sectors.

8.0

2026-10-16

RESEARCH · 1 SOURCE · InfoQ AI, ML & Data Engineering

Google open-sources EnvHarness to adapt training environments for self-evolving agents

Google researchers open-sourced EnvHarness, a framework that wraps static environments via the standard reset()/step() interface and applies composable setup, rule, and link mutations so environments adapt to an agent's weaknesses without changing the underlying environment or evaluators. The release also includes EnvRigger, an automated loop that analyzes agent rollouts to generate and validate targeted environment edits; Google reports consistent performance gains (≈2.9–3.7 percentage points) across benchmarks such as WebArena, SWE-bench Verified, and ALFWorld on models including Gemini, Qwen, and Claude Sonnet.

7.0

2026-09-30

CODING · 1 SOURCE · AWS Machine Learning

AWS shows multi-agent music pipeline on Amazon Bedrock AgentCore Runtime Instances

AWS published a walkthrough demonstrating a three-agent music production pipeline deployed on Amazon Bedrock AgentCore Runtime Instances. The example colocates agents on a single EC2-backed instance with persistent volumes and an NVIDIA L4 GPU (running ACE-Step for on-instance music generation), uses shared runtimeSessionId for agent collaboration, and packages agents as ECR containers or S3 artifacts to enable multi-day, stateful workflows.

6.0

2026-09-17

RESEARCH · 1 SOURCE · arXiv cs.AI

arXiv paper introduces SAFE benchmark to test whether frontier models seek safety evidence before acting

The paper (arXiv:2609.17865v1) introduces SAFE, a controlled benchmark where models decide whether to retrieve optional safety-relevant evidence that varies in cost, probability, severity, and presentation. Evaluating GPT-5.5, o3, Claude Opus 4.8, and Claude Sonnet 4.6, the authors find distinct evidence-acquisition policies (Opus inspects by default, o3 skips most, GPT-5.5 and Sonnet intermediate), strong sensitivity to severity and retrieval cost, weaker sensitivity to probability, and a mismatch between stated rationales and actual causal influences.

7.0