Tatiana Gaintseva — accepted AI-conference papers

2026

CASteer: Cross-Attention Steering for Controllable Concept Erasure

Tatiana Gaintseva, Andreea-Maria Oncescu, Chengcheng Ma, Ziquan Liu, Martin Benning, Gregory Slabaugh +2

ICLR 2026poster

Diffusion models have transformed image generation, yet controlling their outputs for diverse applications, including content moderation and creative customization, remains challenging. Existing approaches usually require task-specific training and struggle to generalise across both concrete (e.g.,…

Cited by 0 Source ↗code ↗Scholar ↗

2026

MIDSTEER: Optimal Affine Framework for Steering Generative Models

Tatiana Gaintseva, Andrew Stepanov, Ziquan Liu, Martin Benning, Gregory Slabaugh, Jiankang Deng +1

ICML 2026poster

Steering intermediate representations has emerged as a powerful strategy for controlling generative models. However, despite its empirical success, it currently lacks a comprehensive theoretical framework. In this paper, we bridge this gap by formalizing the theory of concept steering. First, we est…

Cited by 0 Source ↗Scholar ↗

2026

Multi-Way Representation Alignment

Akshit Achara, Tatiana Gaintseva, Matéo Mahaut, Pritish Chakraborty, Viktor Johansson, Melih Barsbey +2

ICML 2026poster

The Platonic Representation Hypothesis suggests that independently trained neural networks converge to increasingly similar latent spaces. However, current strategies for mapping these representations are inherently pairwise, scaling quadratically with the number of models and failing to yield a con…

Cited by 0 Source ↗Scholar ↗