Related papers: Dialz: A Python Toolkit for Steering Vectors
Large language models (LLMs) can sometimes detect when they are being evaluated and adjust their behavior to appear more aligned, compromising the reliability of safety evaluations. In this paper, we show that adding a steering vector to an…
We introduce VLM-Lens, a toolkit designed to enable systematic benchmarking, analysis, and interpretation of vision-language models (VLMs) by supporting the extraction of intermediate outputs from any layer during the forward pass of…
We present DIALIGHT, a toolkit for developing and evaluating multilingual Task-Oriented Dialogue (ToD) systems which facilitates systematic evaluations and comparisons between ToD systems using fine-tuning of Pretrained Language Models…
Large language models (LLMs) are used in a variety of mission-critical roles. Due to the rapidly developing nature of LLMs, there is a lack of quantifiable understanding of the bias and perspective associated with LLM output. Inspired by…
Large language model (LLM) steering has emerged as a promising paradigm for controlling model behavior at inference time through targeted manipulation of hidden states, offering a lightweight alternative to expensive retraining. However,…
Since Lorenz's seminal work on a simplified weather model, the numerical analysis of nonlinear dynamical systems has become one of the main subjects of research in physics. Despite of that, there remains a need for accessible, efficient,…
We introduce Directional Stimulus Prompting, a novel framework for guiding black-box large language models (LLMs) toward specific desired outputs. Instead of directly adjusting LLMs, our method employs a small tunable policy model (e.g.,…
Causal inference is a statistical paradigm for quantifying causal effects using observational data. It is a complex process, requiring multiple steps, iterations, and collaborations with domain experts. Analysts often rely on visualizations…
Large Language Models (LLMs), despite advances in instruction tuning, often fail to follow complex user instructions. Activation steering techniques aim to mitigate this by manipulating model internals, but have a potential risk of…
Large language models (LLMs) are increasingly deployed as customer-facing agents, yet evaluating their reliability remains challenging due to stochastic, multi-turn interactions. Current evaluation protocols rely on linear Monte Carlo…
As large language models increasingly mediate firm - customer interactions, firms face a tradeoff: the most capable models perform well but are costly and difficult to control at scale. Existing knowledge distillation methods address this…
Steering vectors have emerged as a lightweight and effective approach for aligning large language models (LLMs) at inference time, enabling modulation over model behaviors by shifting LLM representations towards a target behavior. However,…
Pre-trained language models and other generative models have revolutionized NLP and beyond. However, these models tend to reproduce undesirable biases present in their training data. Also, they may overlook patterns that are important but…
In recent years, instruction tuning has gained increasing attention and emerged as a crucial technique to enhance the capabilities of Large Language Models (LLMs). To construct high-quality instruction datasets, many instruction processing…
Steering vectors are a promising approach to aligning language model behavior at inference time. In this paper, we propose a framework to assess the limitations of steering vectors as alignment mechanisms. Using a framework of transformer…
While test-time reasoning enables language models (LMs) to tackle complex tasks, searching or planning in natural language can be slow, costly, and error-prone. But even when LMs struggle to emulate the precise reasoning steps needed to…
We describe how Python can be leveraged to streamline the curation, modelling and dissemination of drug discovery data as well as the development of innovative, freely available tools for the related scientific community. We look at various…
Large Language Models (LLMs) exhibit impressive performance across diverse domains but often suffer from overconfidence, limiting their reliability in critical applications. We propose SteerConf, a novel framework that systematically steers…
Large language models (LLMs) are increasingly being used to generate synthetic datasets for the evaluation and training of downstream models. However, prior work has noted that such generated data lacks diversity. In this paper, we propose…
Large language models (LLMs) increasingly serve as automated evaluators, yet they suffer from "self-preference bias": a tendency to favor their own outputs over those of other models. This bias undermines fairness and reliability in…