AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

PRISM: Enhancing PRotein Inverse Folding through Fine- Grained Retrieval on Structure-Sequence Multimodal Representations

Sazan Mahbub, Souvik Kundu, Eric P. Xing

ICLR 2026 · Probabilistic and scientific modeling › Molecules and proteins · approach: Generative protein structure design · 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
add fine-grained retrieval with hybrid decoderAIDO.Protein-IF · CATH-4.2 test (all)AAR (%): 58.60 -> 60.43Improved
add retrievalAIDO.Protein-IF · CATH-4.2 test (all)perplexity: 2.94 -> 2.71Improved
add retrievalAIDO.Protein-IF · TS500perplexity: 2.42 -> 2.27Mixed
add retrieval with structure-only encoder (ablation)ProteinMPNN-CMLM · CATH-4.2AAR (%): 48.62 -> 58.02Improved
add self-attention to cross-attention aggregation (ablation)PRISM · TS50AAR (%): 64.23 -> 67.92Improved
increase MHSCA blocks from 1 to 2 (ablation)PRISM · CATH-4.2 testAAR (%): 60.23 -> 60.43Improved
increase MHSCA blocks from 2 to 3 (ablation)PRISM · CATH-4.2 testAAR (%): 60.43 -> 60.35No clear effect
finetune joint encoder (ablation)PRISM_pifold · CATH-4.2 testAAR (%): 60.09 -> 62.01Improved
add inference-time iterative refinement (ablation)PRISM_pifold (orig. G) · CATH-4.2 testAAR (%): 60.09 -> 61.51Improved
increase retrieved entries K from 1 to 5 (ablation)PRISM · CATH-4.2 valperplexity: 2.788 -> 2.709Improved
increase K beyond 35 (ablation)PRISM · CATH-4.2 valperplexity: stabilizes around 2.681No clear effect
extend retrieval database with PDB entries (ablation)PRISM · CAMEO 2022AAR (%): ~64.6 unchangedNo clear effect

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