HRM arcprize.org leaderboard

Do you plan to submit your model to the arcagi leaderboard ?

Jul 28 '25 20:07 vavans

It's working in progress now 😃

Jul 29 '25 06:07 imoneoi

Hi..Any update on this?

Aug 10 '25 05:08 narvind2003

https://x.com/GregKamradt/status/1954173118416629760

Aug 10 '25 15:08 davtoro

@davtoro please report back later, since people are accusing HRM of data leaking... but the private test set could help with the whole situation I guess? https://github.com/codelion/openevolve/discussions

Aug 11 '25 06:08 BradKML

https://x.com/arcprize/status/1956431617951740044

Aug 16 '25 00:08 davtoro

https://arcprize.org/blog/hrm-analysis

Aug 16 '25 00:08 davtoro

@davtoro, if it is all just "outer loop" refinement, then does this mean it is similar to universal transformers? (not gonna comment on data augmentation since that is just a self-consistency thing) https://arxiv.org/pdf/1807.03819 Also, why do these two pop into my head? https://arxiv.org/html/2502.05171v2 https://arxiv.org/html/2412.06769 https://www.themoonlight.io/en/review/reasoning-beyond-language-a-comprehensive-survey-on-latent-chain-of-thought-reasoning P.S. Please mix HRM with Diffusion models, can't wait to see how that could pan out https://github.com/Jianguo99/OpenDLM

Aug 16 '25 00:08 BradKML