TasksSotADatasetsPapersMethodsSubmitAbout
Papers With Code 2

A community resource for machine learning research: papers, code, benchmarks, and state-of-the-art results.

Explore

Notable BenchmarksAll SotADatasetsPapersMethods

Community

Submit ResultsAbout

Data sourced from the PWC Archive (CC-BY-SA 4.0). Built by the community, for the community.

Models/MolReFlect

MolReFlect

Reported on 20 benchmarks across 3 tasks

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Medical20 results

  • Drug DiscoveryonChEBI-20
    BLEU
    90.3
    best: 92.6 (LDMol)
  • Drug DiscoveryonChEBI-20
    Exact Match
    51
    best: 53.3 (LDMol)
  • Drug DiscoveryonChEBI-20
    Levenshtein
    11.84
    best: 28.54 (Text+Chem T5 small)
  • Drug DiscoveryonChEBI-20
    MACCS FTS
    92.9
    best: 97.3 (LDMol)
  • Drug DiscoveryonChEBI-20
    Morgan FTS
    81.3
    best: 93.1 (LDMol)
  • Drug DiscoveryonChEBI-20
    RDK FTS
    86
    best: 95 (LDMol)
  • Drug DiscoveryonChEBI-20
    Validity
    97.7
    best: 100 (BioT5+)
  • Molecule CaptioningonChEBI-20
    BLEU-2
    67.6
    best: 73.2 (Mol-LLM (Mistral-Instruct-v0.2))
  • Molecule CaptioningonChEBI-20
    BLEU-4
    60.8
  • Molecule CaptioningonChEBI-20
    METEOR
    68
    best: 68.1 (BioT5+)
  • Molecule CaptioningonChEBI-20
    ROUGE-1
    70.3
    best: 71 (BioT5+)
  • Molecule CaptioningonChEBI-20
    ROUGE-2
    57.1
    best: 61.8 (XMolCap)
  • Molecule CaptioningonChEBI-20
    ROUGE-L
    64.4
    best: 65.3 (PEIT-GEN)
  • Text-based de novo Molecule GenerationonChEBI-20
    BLEU
    90.3
    best: 92.6 (LDMol)
  • Text-based de novo Molecule GenerationonChEBI-20
    Exact Match
    51
    best: 53.3 (LDMol)
  • Text-based de novo Molecule GenerationonChEBI-20
    Levenshtein
    11.84
    best: 28.54 (Text+Chem T5 small)
  • Text-based de novo Molecule GenerationonChEBI-20
    MACCS FTS
    92.9
    best: 97.3 (LDMol)
  • Text-based de novo Molecule GenerationonChEBI-20
    Morgan FTS
    81.3
    best: 93.1 (LDMol)
  • Text-based de novo Molecule GenerationonChEBI-20
    RDK FTS
    86
    best: 95 (LDMol)
  • Text-based de novo Molecule GenerationonChEBI-20
    Validity
    97.7
    best: 100 (BioT5+)