LIVE PULSE
3.9 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.1 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.0 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src1.7 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.4 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.1 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.1 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.1 Study examines issue bias in LLMs used as writing assistants before Swedish 2026 election — 1 src1.1 Study Audits Misalignment in Multi-Modal World Models — 1 src1.1 Retrieval-Grounded Reasoning Approach Proposed for Universal Multimodal Embeddings — 1 src3.9 Anthropic CEO Amodei calls for slower AI development and shared safety rules — 11 src2.1 Agility Robotics unveils Digit 5 humanoid for warehouses and factories — 2 src2.0 Apple ships rebuilt Siri with Google Gemini, but not in the EU — 2 src1.7 Siri AI in macOS 27 Golden Gate: FAQ, Germany availability, privacy questions — 2 src1.4 Sam Altman says OpenAI will not go public in 2026, citing AI safety concerns — 5 src1.1 OpenAI contractors review real ChatGPT conversations to rate responses, report says — 2 src1.1 Anthropic data retention policy prompts firms to limit Claude use for sensitive work — 1 src1.1 Study examines issue bias in LLMs used as writing assistants before Swedish 2026 election — 1 src1.1 Study Audits Misalignment in Multi-Modal World Models — 1 src1.1 Retrieval-Grounded Reasoning Approach Proposed for Universal Multimodal Embeddings — 1 src
HEATPULSEAI MAGAZINES
FLIP · FOLLOW · SAVE
Log inSign up

model retraining

topic3 events
papersSEP 10 04:00 UTC

Study Compares Retraining Policies for Subgroup Disparity Under Data Drift

A new arXiv paper examines how the choice of retraining policy affects subgroup error rates in deployed classifiers as data distributions drift. The authors run paired comparisons of complete scheduled retraining against loss-triggered and subgroup-gap-triggered approaches, tracking cumulative subgroup disparity across model sequences, including gaps between updates. The work frames retraining timing as a question of fairness measurement rather than accuracy alone.

fairness/retraining/distribution-shift/evaluation
papersSEP 10 04:00 UTC

Layer-Selective Unlearning Method Targets Sensitive Content in Large Language Models

Researchers have proposed a machine unlearning technique that directs the removal of memorized information to specific layers of a large language model rather than treating the whole network. Framed as a lighter-weight alternative to full retraining, the approach seeks to erase sensitive, copyrighted, or otherwise undesirable training content while preserving the model's remaining capabilities.

machine-unlearning/llm/privacy/safety
papersSEP 10 04:00 UTC

Paper proposes measuring and optimizing LLM agent harnesses without retraining models

A new arXiv paper studies how LLM tool agents can be improved by modifying the runtime harness around a fixed model, including prompts, tool interfaces, middleware, state handling, and recovery logic. The authors frame this as a resource-bounded harness selection problem, arguing that agent performance can be improved without retraining. The work offers ways to measure and optimize these harness components systematically.

agents/evaluation/benchmarks