Cross-modal steering vectors transfer zero-shot between text and image generation
benno_krojer · x · 2026-10-28
A new preprint by Benno Krojer and Donghao Zeng asks how multimodal unified multimodal models really are. They show that steering vectors learned in one modality transfer zero-shot to the other, in both directions (text ↔ image generation). Co-author Tamar Rott Shaham highlights this as evidence for shared causal directions — and a shared conceptual space — across modalities.
More from Models
- Cognition's FrontierCode 2.0 chart accused of cherry-picking after DeepSeek score gaps surface — teortaxesTex · 2026-10-29
- Google ships Gemini 4 Carbon two weeks after Argon, claims SOTA on cyber and coding — tulseedoshi · 2026-10-29
- Claude's safeguards poison entire chat over benign request; user had to surgically edit MEMORY.md to revive it — RexDouglass · 2026-10-29
- Early Take on Gemini 4 Carbon: Lots of Thinking, Good Model Overall — SamuelAlbanie · 2026-10-29
- Gemini 4 Carbon tops AA Index v5, beats Opus 5.5 and GPT-6 Astra on most benchmarks — Yuchenj_UW · 2026-10-29
- Gemini 4 Carbon Keeps $2/$10 Pricing, Scores 59 on AA v5.0 — _philschmid · 2026-10-29