Controlling Language and Diffusion Models by Transporting Activations

• URL: http://arxiv.org/abs/2410.23054v2
• Date: Fri, 22 Nov 2024 16:04:44 GMT
• Title: Controlling Language and Diffusion Models by Transporting Activations
• Authors: Pau Rodriguez, Arno Blaas, Michal Klein, Luca Zappella, Nicholas Apostoloff, Marco Cuturi, Xavier Suau,
• Abstract summary: We introduce Activation Transport (AcT), a framework to steer activations guided by optimal transport theory. We experimentally show the effectiveness and versatility of our approach by addressing key challenges in large language models (LLMs) and text-to-image diffusion models (T2Is)
• Score: 23.352500740697938
• License:
• Abstract: The increasing capabilities of large generative models and their ever more widespread deployment have raised concerns about their reliability, safety, and potential misuse. To address these issues, recent works have proposed to control model generation by steering model activations in order to effectively induce or prevent the emergence of concepts or behaviors in the generated output. In this paper we introduce Activation Transport (AcT), a general framework to steer activations guided by optimal transport theory that generalizes many previous activation-steering works. AcT is modality-agnostic and provides fine-grained control over the model behavior with negligible computational overhead, while minimally impacting model abilities. We experimentally show the effectiveness and versatility of our approach by addressing key challenges in large language models (LLMs) and text-to-image diffusion models (T2Is). For LLMs, we show that AcT can effectively mitigate toxicity, induce arbitrary concepts, and increase their truthfulness. In T2Is, we show how AcT enables fine-grained style control and concept negation.

Related papers

• CAR: Controllable Autoregressive Modeling for Visual Generation [100.33455832783416]
  Controllable AutoRegressive Modeling (CAR) is a novel, plug-and-play framework that integrates conditional control into multi-scale latent variable modeling. CAR progressively refines and captures control representations, which are injected into each autoregressive step of the pre-trained model to guide the generation process. Our approach demonstrates excellent controllability across various types of conditions and delivers higher image quality compared to previous methods.
  arXiv  Detail & Related papers  (2024-10-07T00:55:42Z)
• Unleashing the Power of Task-Specific Directions in Parameter Efficient Fine-tuning [65.31677646659895]
  This paper focuses on the concept of task-specific directions (TSDs)-critical for transitioning large models from pretrained states to task-specific enhancements in PEFT. We introduce a novel approach, LoRA-Dash, which aims to maximize the impact of TSDs during the fine-tuning process, thereby enhancing model performance on targeted tasks.
  arXiv  Detail & Related papers  (2024-09-02T08:10:51Z)
• Towards Human-Like Driving: Active Inference in Autonomous Vehicle Control [0.5437298646956507]
  This paper presents a novel approach to Autonomous Vehicle (AV) control through the application of active inference. Active inference is a theory derived from neuroscience that conceptualizes the brain as a predictive machine. Our method integrates active inference with deep learning to manage lateral control in AVs, enabling them to perform lane following maneuvers within a simulated urban environment.
  arXiv  Detail & Related papers  (2024-07-10T14:08:27Z)
• Steering Llama 2 via Contrastive Activation Addition [41.54815073311959]
  Contrastive Activation Addition (CAA) is a method for steering language models by modifying their activations during forward passes. CAA accurately steers model outputs and sheds light on how high-level concepts are represented in Large Language Models (LLMs)
  arXiv  Detail & Related papers  (2023-12-09T04:40:46Z)
• Learning Exactly Linearizable Deep Dynamics Models [0.07366405857677226]
  We propose a learning method for exactly linearizable dynamical models that can easily apply various control theories to ensure stability, reliability, etc. The proposed model is employed for the real-time control of an automotive engine, and the results demonstrate good predictive performance and stable control under constraints.
  arXiv  Detail & Related papers  (2023-11-30T05:40:55Z)
• Exploring Model Transferability through the Lens of Potential Energy [78.60851825944212]
  Transfer learning has become crucial in computer vision tasks due to the vast availability of pre-trained deep learning models. Existing methods for measuring the transferability of pre-trained models rely on statistical correlations between encoded static features and task labels. We present an insightful physics-inspired approach named PED to address these challenges.
  arXiv  Detail & Related papers  (2023-08-29T07:15:57Z)
• Model-Based Reinforcement Learning with Isolated Imaginations [61.67183143982074]
  We propose Iso-Dream++, a model-based reinforcement learning approach. We perform policy optimization based on the decoupled latent imaginations. This enables long-horizon visuomotor control tasks to benefit from isolating mixed dynamics sources in the wild.
  arXiv  Detail & Related papers  (2023-03-27T02:55:56Z)
• Guided Conditional Diffusion for Controllable Traffic Simulation [42.198185904248994]
  Controllable and realistic traffic simulation is critical for developing and verifying autonomous vehicles. Data-driven approaches generate realistic and human-like behaviors, improving transfer from simulated to real-world traffic. We develop a conditional diffusion model for controllable traffic generation (CTG) that allows users to control desired properties of trajectories at test time.
  arXiv  Detail & Related papers  (2022-10-31T14:44:59Z)
• Your Autoregressive Generative Model Can be Better If You Treat It as an Energy-Based One [83.5162421521224]
  We propose a unique method termed E-ARM for training autoregressive generative models. E-ARM takes advantage of a well-designed energy-based learning objective. We show that E-ARM can be trained efficiently and is capable of alleviating the exposure bias problem.
  arXiv  Detail & Related papers  (2022-06-26T10:58:41Z)
• Isolating and Leveraging Controllable and Noncontrollable Visual Dynamics in World Models [65.97707691164558]
  We present Iso-Dream, which improves the Dream-to-Control framework in two aspects. First, by optimizing inverse dynamics, we encourage world model to learn controllable and noncontrollable sources. Second, we optimize the behavior of the agent on the decoupled latent imaginations of the world model.
  arXiv  Detail & Related papers  (2022-05-27T08:07:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.