AI Paper Atlas

Research map / Large language models / Instruction and fine-tuning

Entropy-Guided Dynamic Tokens for Graph-LLM Alignment in Molecular Understanding

Zihao Jing, QIUHAO Zeng, Ruiyi Fang, Yan Sun, Boyu Wang, Pingzhao Hu

ICLR 2026 · Large language models › Instruction and fine-tuning · approach: Calibration and knowledge in LLM tuning · 1 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
add entropy-guided patching instead of BRICS fragments (ablation)EDT-Former · Pampaaccuracy (%): 71.90 -> 84.52Improved
add entropy-guided patching instead of random patches (ablation)EDT-Former · BBBPaccuracy (%): 68.90 -> 75.06Improved
add any patching (entropy) versus none (ablation)EDT-Former · BBBPaccuracy (%): 39.67 -> 75.06Improved
add entropy-guided patching versus BRICS on BBBP (ablation)EDT-Former · BBBPaccuracy (%): 73.59 -> 75.06Improved
add modality fusion (text, graph, 3D alignment) to the connector (ablation)EDT-Former · MoleculeQAaverage accuracy: 26% avg. degradation when removedImproved
add Dynamic Query Transformer (dynamic tokens) instead of fixed-length projection (ablation)EDT-Former · MoleculeQAaverage accuracy: 10.7% average drop when removedImproved
add Entropy-Guided Patching instead of uniform 3-token SMILES patches (ablation)EDT-Former · MoleculeQAaverage accuracy: 11.5% average drop when removedImproved
add the 3D modality to text and graph (ablation)EDT-Former · MoleculeQAtotal accuracy (%): 65.11 -> 68.34Improved
add the graph modality to text and 3D (ablation)EDT-Former · MoleculeQAtotal accuracy (%): 60.48 -> 68.34Improved
add graph and 3D inputs to text-only SMILES (ablation)EDT-Former · MoleculeQAtotal accuracy (%): 58.77 -> 68.34Improved
attach EDT-Former connector to Mistral-8BMistral-8B · MoleculeQAtotal accuracy (%): 28.00 -> 55.63Improved
attach EDT-Former connector to Llama3.1-8BLlama3.1-8B · MoleculeQAtotal accuracy (%): 49.31 -> 68.34Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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