AI FOR SCIENCE · ON THE RECORD
Carnegie Mellon University
On the timeline →People filter →
Research on the record
- 2026-10-13Nine agents write 135 follow-up papers to order for national-lab scientists; a judge finds 114 of them fail an integrity checkLevel 2 · Computed
- 2026-09-23WhatWorkedBench tests whether AI research agents know why an experiment worked, not just how to run itLevel 2 · Computed
- 2026-01-23Smirk shows common molecular tokenizers silently erase rare chemistry, and closes the gap with 165 tokensLevel 2 · Computed
- 2025-08-05ODAC25, a 60-million-calculation dataset for direct-air-capture sorbents, takes 460 of its MOFs from CuspAI's generatorLevel 3 · Checked against existing evidence
- 2025-06-24AIMNet2-rxn, a machine-learned potential trained on reacting molecules, puts organic reaction barriers within about 2 kcal/mol of the quantum chemistry it learned fromLevel 2 · Computed
- 2023-12-20Coscientist plans and executes chemistry experimentsLevel 4 · Tested in the physical world
Graded by how far the evidence went, not how good the result is. See the methodology.
Read on AI4S.fyi
People
Only people the AI4S.fyi record has named: in a round, a timeline entry, commentary, an interview or a sourced role. Not the full team.
Named on AI4S.fyi
Jeremy AvigadDean's Chair in Logic and Philosophy of Mathematics, Departments of Philosophy and Mathematical Sciences How to Formalize All of Mathematics- SSean WelleckAssistant Professor, Language Technologies Institute, School of Computer Science Nine agents write 135 follow-up papers to order for national-lab scientists; a judge finds 114 of them fail an integrity check
- SSeungone KimPhD student, Language Technologies Institute Nine agents write 135 follow-up papers to order for national-lab scientists; a judge finds 114 of them fail an integrity check
- XXueqi LiAuthor affiliation on a 2026 paper WhatWorkedBench tests whether AI research agents know why an experiment worked, not just how to run it
Yibo KongStudent, Master of Computational Data Science WhatWorkedBench tests whether AI research agents know why an experiment worked, not just how to run it
Po-Shen LohProfessor Why Do We Need Human Mathematicians Anymore?
Eric XingProfessor, Machine Learning Department GenBio AI, co-founded by a Nobel laureate, releases a preview of AIDO Cell, a model that simulates a human cell from DNA to whole-cell behaviour
Olexandr IsayevCarl and Amy Jones Professor in Interdisciplinary Science, Chemistry AIMNet2-rxn, a machine-learned potential trained on reacting molecules, puts organic reaction barriers within about 2 kcal/mol of the quantum chemistry it learned from
This page compiles what AI4S.fyi has recorded and links each item to its source. It is not a grade of the organization and not its complete history: a blank means the record has not caught up as often as it means nothing happened.