Domain on the map

AI science benchmarks and RL environments for science

Scientific research work, such as handling lab and biological data, reading the literature and reasoning about experiments.

RL environments
7
AI benchmarks
70
Shares entries with
18 categories

Snapshot 2026-11-05. How categories are built · Where the data comes from

The galaxy could not load. See every vendor on the board.

Science, live from the universe. Hover a star for its name; click to open it. Open it in the universe

In short

What are RL environments for science?
A simulated workplace where an AI agent practises a task (a code repository, a browser, business software) and an automatic grader scores each attempt. Labs train on thousands of these attempts. Bought to improve models. For science, that means scientific research work, such as handling lab and biological data, reading the literature and reasoning about experiments.
Who builds RL environments for science?
Surge AI, Latch, Emulated, ReasonCore, Ulam and Rise Data Labs, among 7 confirmed below.
What are the main science benchmarks for LLMs?
ProteinGym, BioMysteryBench Human-Difficult, BioMysteryBench Human-Solvable, BixBench, FormationEval and SWE, among 65 confirmed below.

RL environments 7

  1. Surge AI
  2. Latch
  3. Emulated
  4. ReasonCore
  5. Ulam
  6. Rise Data Labs
  7. Tacit Labs

AI benchmarks 70

  1. ProteinGym
  2. BioMysteryBench Human-DifficultAnthropic
  3. BioMysteryBench Human-SolvableAnthropic
  4. BixBench
  5. FormationEvalAlmaz Ermilov
  6. SWEEmulated
  7. Terminal-Bench-Science
  8. Math RL Tasks
  9. ChemBenchLab for AI for Materials, University of Jena
  10. scBenchLatchBio
  11. SciKnowEvalHICAI-ZJU
  12. SpatialBench
  13. TDC AMESTherapeutics Data Commons, Harvard, mims-harvard
  14. TDC BBB_MartinsTherapeutics Data Commons
  15. TDC Bioavailability_MaTherapeutics Data Commons
  16. TDC Caco2_WangTherapeutics Data Commons, Harvard, mims-harvard
  17. TDC Clearance_Hepatocyte_AZTherapeutics Data Commons
  18. TDC Clearance_Microsome_AZTherapeutics Data Commons, Harvard, mims-harvard
  19. TDC CYP2C9_Substrate_CarbonMangelsTherapeutics Data Commons
  20. TDC CYP2C9_VeithTherapeutics Data Commons
  21. TDC CYP2D6_Substrate_CarbonMangelsTherapeutics Data Commons, Harvard, mims-harvard
  22. TDC CYP2D6_VeithTherapeutics Data Commons, Harvard, mims-harvard
  23. TDC CYP3A4_Substrate_CarbonMangelsTherapeutics Data Commons
  24. TDC CYP3A4_VeithTherapeutics Data Commons
  25. TDC DILITherapeutics Data Commons
  26. TDC DRD3 DockingTherapeutics Data Commons, Harvard, mims-harvard
  27. TDC Half_Life_ObachTherapeutics Data Commons
  28. TDC hERGTherapeutics Data Commons, mims-harvard, Zitnik lab at Harvard
  29. TDC HIA_HouTherapeutics Data Commons, Harvard, mims-harvard
  30. TDC LD50_ZhuTherapeutics Data Commons
  31. TDC Lipophilicity_AstraZenecaTherapeutics Data Commons, Harvard, mims-harvard
  32. TDC Pgp_BroccatelliTherapeutics Data Commons, Harvard, mims-harvard
  33. TDC PPBR_AZTherapeutics Data Commons
  34. TDC Solubility_AqSolDBTherapeutics Data Commons, Harvard, mims-harvard
  35. TDC VDss_LombardoTherapeutics Data Commons, mims-harvard, Harvard
  36. VariantBenchLatchBio
  37. GeneBenchOpenAI
  38. PhysicsFinals
  39. PROBE Binding Affinity
  40. PROBE Function Prediction
  41. PhysicsMind
  42. A^3-BenchXi'an Jiaotong University, Beijing Normal University, National University of Singapore
  43. ActiveSciBench
  44. AgroBenchUniversity of Osaka, AIST, Kyoto University, Risa Shinoda et al
  45. FrontierChallengeApodex AI
  46. SCICONVBENCH
  47. TDC BindingDB Patent DTI-DGTherapeutics Data Commons, Harvard Zitnik lab, mims-harvard
  48. TDC TCHard TCR-EpitopeTherapeutics Data Commons, MIMS lab, Harvard
  49. CUSP
  50. Geo-Eval
  51. GSO-Bench
  52. gwBenchmarks
  53. MaCBenchLab for AI for Materials, M3RG
  54. MLReplicate
  55. NatureBenchFrontisAI
  56. Proof of TimeAIM Harvard
  57. SoundnessBenchUniversity of Maryland, College Park
  58. LAB-BenchFutureHouse
  59. PDEBenchTakamoto, Praditia, NEC Laboratories Europe, University of Stuttgart
  60. Collider-Bench
  61. EngiBench
  62. MoleculeNet
  63. OpenADMET ExpansionRx
  64. Prediction Bottlenecks Don't Discover Causal Structure
  65. SciBench
5 not yet confirmedListed in the catalog; no line from its own site or paper confirms it yet.
  1. LifeSciBench
  2. Autoresearch
  3. benchmarks.bio
  4. PROBE Family Classification
  5. PROBE Semantic Similarity

Data 14

  1. Hillclimbbased here
  2. Abundantalso covers it
  3. Aviroalso covers it
  4. Cuaalso covers it
  5. Emulatedalso covers it
  6. Mercoralso covers it
  7. Paretoalso covers it
  8. Patronus AIalso covers it
  9. ReasonCorealso covers it
  10. Reveloalso covers it
  11. Symbalalso covers it
  12. Turingalso covers it
  13. Ulamalso covers it
  14. Vetto AIalso covers it

Next: pick another domain. Tags come from the RL Research daily scrape, canonicalized nightly.