Skip to content

Add FLAS to leaderboard#152

Open
Lunamos wants to merge 1 commit into
stanfordnlp:mainfrom
Lunamos:main
Open

Add FLAS to leaderboard#152
Lunamos wants to merge 1 commit into
stanfordnlp:mainfrom
Lunamos:main

Conversation

@Lunamos

@Lunamos Lunamos commented Jul 3, 2026

Copy link
Copy Markdown

Summary

  • Add a "Beyond rank-1" subsection under the rank-1 steering leaderboard in the README, with FLAS (Flow-based Activation Steering) as a first entry

  • Why a separate section: FLAS is not a rank-1 intervention. Rather than computing one steering vector per prompt and adding it at every position (as HyperSteer does), it integrates a learned velocity field over multiple steps, where the displacement is re-evaluated from the current activation at each step and differs across token positions. Recent methods like AcT and K-Steering also relax these restrictions, so this class may keep growing.

  • Scores follow the main table's setting (held-in harmonic mean, layer 20, Concept16k): 1.025 on Gemma-2-2B-IT, 1.185 on Gemma-2-9B-IT, at a single fixed flow time with no per-concept tuning

  • Left index.html untouched for now — happy to mirror the section there too if you'd like

  • Website: https://flas-ai.github.io

  • Paper: https://arxiv.org/abs/2605.05892

  • Code: https://github.com/flas-ai/FLAS

  • Models: https://huggingface.co/flas-ai

Possible follow-ups (if there's interest)

  • Train a rank-1-constrained FLAS-r1 for the main table, to see how much of the gain survives the restriction

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant