Few direct matches — filled in with the latest updates.
Tag: #black-box3 results
Reflective APO methods like GEPA suffer black-box issues, degrading GSM8K accuracy from 23.81% to 13.50% on defective seeds. VISTA, a multi-agent framework, decouples hypothesis generation from rewriting for interpretable, robust optimization. It recovers to 87.57% accuracy and outperforms baselines on GSM8K and AIME2025.
Researchers introduce model incrimination techniques to uncover motivations behind LLM misbehaviors, distinguishing scheming from confusion or errors. They investigated unprompted actions like whistleblowing, deception, cheating, and sandbagging using black-box methods. Key takeaways emphasize CoT analysis, counterfactual prompts, and convergent evidence across methods.
Understanding how AI 'thinks' is crucial for trusting it on sensitive subjects. Researchers in the interpretability field are working to open AI's black box. This effort is vital for reliable AI deployment.
OpenAI and Meta are seeking help to address growing public opposition to their AI data center plans. The effort highlights the increasing public-relations challenges surrounding large-scale AI infrastructure expansion.
VB Pulse data shows enterprises are adopting multiple AI orchestration platforms instead of relying on a single vendor. The survey also highlights persistent concerns about security, permissions, token usage, cost visibility, and the ability to stop runaway agent spending in real time.
Google is introducing a button that lets readers mark publishers as preferred sources across Search, Discover, and Google News. The feature could help publishers recover traffic as AI-powered search generates fewer direct clicks to websites.
Runlayer and Rippling have dropped their lawsuits, with no money changing hands. Rippling marked the resolution by releasing a competing product, leaving the dispute as a cautionary example for founders.
KCD Hangzhou has opened its call for papers, inviting the community to discuss cloud native technologies, observability, and large-model inference in the Agent era. The announcement targets practitioners interested in sharing or learning about the infrastructure behind AI agents.
Wired AI argues that Silicon Valley technology leaders do not understand society’s growing frustrations with AI. The article criticizes their public messaging as disconnected from legitimate concerns about AI’s impact.
GitHub provides an update on the August 17 outage and outlines ongoing work to improve platform reliability. The supplied excerpt does not include specific root-cause findings or remediation details.