AI Paper Atlas

Research map / Large language models / Agents, code and math

SciNav: A Principled Agent Framework for Scientific Coding Tasks

TIANSHU ZHANG, Huan Sun

ICLR 2026 · Large language models › Agents, code and math · approach: Long-horizon LLM agents

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use SciNav TKCTS instead of Self-DebugGPT-4o (2024-11-20) · ScienceAgentBenchSR: 15.0 -> 18.6Improved
use SciNav instead of OpenHandsGPT-4o (2024-05-13) · ScienceAgentBenchSR: 13.1 -> 16.1Improved
use SciNav instead of Self-DebugClaude-3.7 · ScienceAgentBenchSR: 22.5 -> 25.5Improved
use SciNav instead of Self-DebugClaude-3.7 · ScienceAgentBenchVER: 84.3 -> 72.5Hurt
use SciNav instead of Self-DebugGPT-4o (2024-05-13) · ScienceAgentBenchVER: 71.2 -> 66.0Hurt
use relative judgments instead of random selection (ablation)SciNav · ScienceAgentBenchSR: 15.2 -> 18.6Improved
use relative judgments instead of LLM absolute scoring (ablation)SciNav · ScienceAgentBenchSR: 16.2 -> 18.6Improved
increase initial solutions from 1 to 5 (ablation)SciNav · ScienceAgentBench (40 solved tasks)success rate: 40.5 -> 45.2Improved
increase initial solutions from 1 to 2 (ablation)SciNav · ScienceAgentBench (40 solved tasks)success rate: 40.5 -> 40.5No clear effect
add self-improvement (ablation)SciNav · ScienceAgentBench (40 solved tasks)success rate: 45.2 -> 57.1Improved
use K=2 instead of larger K in top-K (ablation)SciNav · DA-Code (30 task preliminary)average scoreImproved
use Claude-3.7 or mixture for solution generation (ablation)SciNav · ScienceAgentBenchSR: 18.6 -> 19.6No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy