Teresa Head-Gordon Lab retweeted
🎉 Congrats to OpenAI on GPT-5.6's strong performance on Agents' Last Exam!
We created Agents' Last Exam (ALE) to provide realistic, reproducible evaluations and continuous public measurements of frontier AI agents on economically valuable, real-world work across broad domains. ALE currently includes 1,500+ expert-sourced, long-horizon tasks spanning 55 non-physical occupations, contributed by 300+ experts from more than 100 institutions. All tasks are grounded in real professional work and evaluated with fully verifiable, outcome-based grading.
It's exciting to see ALE becoming an important guidepost for frontier agent development—helping the community track progress, better understand emerging capabilities, and build AI systems that are increasingly reliable, capable, and useful in the real world.
A huge thank you to the hundreds of experts and collaborators who have contributed to ALE. This is truly a community effort. We're continuing to expand the benchmark with new tasks and domains, and we'd love for more experts to contribute. Together, we can build an even stronger foundation for measuring progress toward capable, trustworthy Agentic AI.
Submit your contribution: agents-last-exam.org/
Replying to @OpenAI
On Agents' Last Exam, GPT‑5.6 Sol sets a new high of 53.6, eclipsing Claude Fable 5 (adaptive) by 13.1 points.
At medium reasoning, it beats Fable 5 by 11.4 points at roughly one-quarter the estimated cost. GPT‑5.6 Terra and Luna also outperforms Fable 5 at around one-sixteenth the cost.
Why use BFGS when you can have the full Hessian? Transition state refinement needs Hessian info—hard to get with DFT, but not with ML potentials! Just take the second derivative, no special training required. More details: doi.org/10.1038/s41467-024-5…
Teresa Head-Gordon Lab retweeted
Fresh out of the arxiv! In this work, we (Joe, @dyushag , Dorian, and myself @thg_lab ) used supervised fine tuning (SFT) and direct preference optimization (DPO) to modify Llama 3.1 into a chemical language model. See more details here: arxiv.org/abs/2409.02231 (1/4)
🎙️ Our very own Dr. THG dives into "The Ethics of Emerging Technology: The Era of Artificial Intelligence" on this insightful podcast episode. Explore the critical role of ethics in AI with us!
🔗 buzzsprout.com/1204337/15285…
#ArtificialIntelligence #EthicsInTech
Teresa Head-Gordon Lab retweeted
Organized together with @thg_lab @GiEmmePi1 @GroupParrinello @mhg_group & Vanda Glezakou,
with the support of @cecamEvents @IITalk and sponsored by CASALE (casale.ch)
We are delighted to host our lab alumni @Mejohnson81 for the physical chemistry seminar @UCB_Chemistry. We really enjoyed the lunch and the talk! Hope to see your new research progress soon.
Teresa Head-Gordon Lab retweeted
Farewell #EuroQSAR selfie with @fiona_chembot and warmest congrats to her for winning the best prize award🏆
From the South Side of Chicago to Berkeley, California with Stops in Between. An Autobiography of a true legend @UCB_Chemistry. Quoting my favorite sentence: "Advice ... is to be understood and appreciated, but you must make the final judgments..." pubs.acs.org/doi/10.1021/acs…
Teresa Head-Gordon Lab retweeted
Many experimentally known high-pressure ice phases are structurally very similar. In a new paper
nature.com/articles/s41467-0…
we combine a ML potential and statistical mechanics to reveal the nature of phase boundaries and suggest novel solid-solid transition mechanisms.
Teresa Head-Gordon Lab retweeted
Today we're pleased to share this infographic NewtonNet by the @thg_lab. Check out the full paper #openaccess here: doi.org/10.1039/D2DD00008C #compchem #machinelearning
Teresa Head-Gordon Lab retweeted
Hello new home for the next few years 👋 #Berkeley
Teresa Head-Gordon Lab retweeted
If you're following this ongoing back and forth happening in the chemical literature, the latest rebuttal to the below has just been published in @JPhysChem Letters
pubs.acs.org/doi/10.1021/acs…
Wonderful to see, given that some people are so worried about censorship and cancel culture in chemistry, that certain concerned parties have recently managed to publish very similar takes in IUPAC (degruyter.com/document/doi/1…) and GDCh publications (onlinelibrary.wiley.com/doi/…)
Teresa Head-Gordon Lab retweeted
Former MolSSI Software Fellow @chenru_duan just successfully defended his PhD @MIT under the direction of Prof. Heather Kulik. We congratulate him on this accomplishment and wish him every success as he begins his professional career!
Teresa Head-Gordon Lab retweeted
“basic research is the foundation on which all technologic advances rest” nature.com/articles/s42254-0…
Teresa Head-Gordon Lab retweeted
Although this is firmly grounded in the context of the physical sciences, the conceptual analysis around, e.g. renaming buildings, is excellent and applicable to many domains.
Led by John Herbert, I am part of a group of scientists that is responding to the recent work of Krylov et al. on the subject of academic "free speech." Check our position, and if you agree with it, please RT to amplify #RealTimeChem #inclusion #respect pubs.acs.org/doi/10.1021/acs…
Teresa Head-Gordon Lab retweeted
Thanks for contributing to these efforts. I was really disappointed to attend a Telluride workshop last month at which Krylov and others were offered another platform to share their regressive and harmful talking points. "Free speech" is not the same as consequence-free speech
Led by John Herbert, I am part of a group of scientists that is responding to the recent work of Krylov et al. on the subject of academic "free speech." Check our position, and if you agree with it, please RT to amplify #RealTimeChem #inclusion #respect pubs.acs.org/doi/10.1021/acs…
Teresa Head-Gordon Lab retweeted
What do we value as a community? I’ve joined John Herbert, Martin Head-Gordon, Teresa Head-Gordon, Rommie Amaro, Alán Aspuru-Guzik, Roald Hoffmann, Carol Parish, Christina Payne, & Troy Van Voorhis in @JPhysChem Letters to share our view. #compchem | pubs.acs.org/doi/10.1021/acs…

























