Skip to content
View maxr0ssi's full-sized avatar
3 followers · 1 following

Achievements

Achievement: QuickdrawAchievement: Pull Sharkx2Achievement: YOLOAchievement: Pair Extraordinaire

Achievements

Achievement: QuickdrawAchievement: Pull Sharkx2Achievement: YOLOAchievement: Pair Extraordinaire

Block or report maxr0ssi

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
maxr0ssi/README.md

Max Rossi

MS Data Science & Research @ Columbia, with the Advanced Master's Research extension.

I study how language models communicate through hidden states, drawing on information theory.

  • CacheBack / RCLC — agent communication in latent space, guided by what the receiver needs. The receiver sends a small internal query to the sending agents, which use it to select what to send back. Paper · Code · Website & demos

  • Latent Cache Flow — model-to-model communication without text. Co-first-author paper at the ICML 2026 Adapt-FM workshop. Paper

Website · LinkedIn

Pinned Loading

  1. agentcacheback/cacheback agentcacheback/cacheback Public

    CacheBack: receiver-conditioned latent communication between agents

    Python 9