logo
today local_bar
Weak-to-strong generalization

4 papers across 3 sessions

Poster Session 3

1 paper
Thursday, December 4, 2025 · 11:00 AM → 2:00 PM
Exhibit Hall C,D,E
Robust SuperAlignment: Weak-to-Strong Robustness Generalization for Vision-Language Models
#5513 Spotlight · Junhao Dong, Cong Zhang, Xinghua Qu, Zejun MA, Piotr Koniusz, Yew Soon Ong

We propose the first robust weak-to-strong generalization framework to elicit robust knowledge from a strong student VLM in an unsupervised scheme.

Poster Session 5

2 papers
Friday, December 5, 2025 · 11:00 AM → 2:00 PM
Exhibit Hall C,D,E
From Linear to Nonlinear: Provable Weak-to-Strong Generalization through Feature Learning
#4003 · Junsoo Oh, Jerry Song, Chulhee Yun

We theoretically investigate weak-to-strong generalization from a linear CNN to a two-layer ReLU CNN
On the Mechanisms of Weak-to-Strong Generalization: A Theoretical Perspective
#3305 · Behrad Moniri, Hamed Hassani

Poster Session 6

1 paper
Friday, December 5, 2025 · 4:30 PM → 7:30 PM
Exhibit Hall C,D,E
Weak-to-Strong Generalization under Distribution Shifts
#2802 · Myeongho Jeon, Jan Sobotka, Suhwan Choi, Maria Brbic

This paper introduces a novel scenario, weak-to-strong generalization under distribution shifts, and proposes Robust AdaptiVe wEightiNg (RAVEN) to tackle this challenge.