AI Paper Atlas

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

Multi-View Encoders for Performance Prediction in LLM-Based Agentic Workflows

Patara Trirat, Wonyong Jeong, Sung Ju Hwang

ICLR 2026 · Large language models › Agents, code and math · approach: LLMs for optimization and heuristics · 3 citations (OpenAlex)

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 multi-view (code+graph+text) encoder with pretraining-free supervised training vs best GNN baselineworkflow performance predictor · FLORA-Bench averageaccuracy: 78.36 -> 79.97Improved
same, utilityworkflow performance predictor · FLORA-Bench averageutility: 73.54 -> 76.33Improved
Agentic Predictor vs GCN on reasoning utilityworkflow performance predictor · FLORA-Bench ReasonAFutility: 91.82 -> 91.47Hurt
add code view to graph+text views (ablation)Agentic Predictor · AFlow averageaccuracy: 75.87 -> 84.38Improved
add text view to code+graph views (ablation)Agentic Predictor · AFlow averageaccuracy: 82.15 -> 84.38Improved
add graph view to code+text views (ablation)Agentic Predictor · AFlow averageaccuracy: 70.82 -> 84.38Improved
graph-only vs full three views on reasoning utility (ablation)Agentic Predictor · AFlow Common Reasoningutility: 91.51 -> 91.47Hurt
use multiple graphs instead of single graph (ablation)Agentic Predictor · AFlow code generationaccuracy: 82.58 -> 84.44Improved
use multiple graphs instead of single graph (ablation)Agentic Predictor · AFlow code generationutility: 78.52 -> 77.22Hurt
add cross-domain unsupervised pretraining at low label ratioAgentic Predictor+ · label ratio 0.1accuracy: ~70 -> >73Improved
use graph predictor instead of 5-shot LLM predictorAgentic Predictor · FLORA-Benchaccuracy: 62.86 -> 84.97Improved
train on AFlow and test on G-Designer (OOD)Agentic Predictor · cross-system OODaccuracy: 59.52 -> 62.05Improved

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