LLM Digest

Story

arxiv_cs_cl · Sep 15, 2026 · paper

Source brief

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

arxiv.orgSep 15, 2026
original source linked

In brief

We introduce and release ScienceBuddy, an interactive scientific research workspace that brings continually improving scientific agents into researchers' everyday workflows. ScienceBuddy supports researchers in carryi...

Feed lens

agentharnessevaluation

Read the original at arxiv.org →Open in live feed Read that day’s brief

ScienceBuddy: Recursive-in-Recursive Self-Improvement for Interactive Scientific Agents

Earlier in this thread 4 items

NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness

TruthInsightBench: An Evidence-Grounded Benchmark for Automated Evaluation of Open-Ended Scientific Discovery Agents

KC-Bench: A Dynamic Interactive Benchmark for Evaluating Knowledge Conflicts in LLM Agents

Recursive Self-Improvement