Exploring Sparse Adapters for Scalable Merging of Parameter Efficient Experts
Samin Yeasar Arnob
Zhan Su
Minseon Kim
Oleksiy Ostapenko
Lucas Caccia
Merging parameter-efficient task experts has recently gained growing attention as a way to build modular architectures that can be rapidly a… (voir plus)dapted on the fly for specific downstream tasks, without requiring additional fine-tuning. Typically, LoRA (Low-Rank Adaptation) serves as the foundational building block of such parameter-efficient modular architectures, leveraging low-rank weight structures to reduce the number of trainable parameters. In this paper, we study the properties of sparse adapters, which train only a subset of weights in the base neural network, as potential building blocks of modular architectures. First, we propose a simple method for training highly effective sparse adapters, which is conceptually simpler than existing methods in the literature and surprisingly outperforms both LoRA and full fine-tuning in our setting. Next, we investigate the merging properties of these sparse adapters by merging adapters for up to 20 natural language processing tasks, thus scaling beyond what is usually studied in the literature. Our findings demonstrate that sparse adapters yield superior in-distribution performance post-merging compared to LoRA or full model merging. Achieving strong held-out performance remains a challenge for all methods considered.
From Intuition to Understanding: Using AI Peers to Overcome Physics Misconceptions
Ruben Weijers
Denton Wu
Hannah Betts
Tamara Jacod
Yuxiang Guan
Vidya Sujaya
Kushal Dev
Toshali Goel
William Delooze
Ying Wu
Kellin Pelrine
Generative AI has the potential to transform personalization and accessibility of education. However, it raises serious concerns about accur… (voir plus)acy and helping students become independent critical thinkers. In this study, we designed a helpful yet fallible AI "Peer" to help students correct fundamental physics misconceptions related to Newtonian mechanic concepts. In contrast to approaches that seek near-perfect accuracy to create an authoritative AI tutor or teacher, we directly inform students that this AI can answer up to 40\% of questions incorrectly. In a randomized controlled trial with 165 students, those who engaged in targeted dialogue with the AI Peer achieved post-test scores that were, on average, 10.5 percentage points higher—with over 20 percentage points higher normalized gain—than a control group that discussed physics history. Qualitative feedback indicated that 91% of the treatment group's AI interactions were rated as helpful. Furthermore, by comparing student performance on pre- and post-test questions about the same concept, along with experts' annotations of the AI interactions, we find initial evidence suggesting the improvement in performance does not depend on the correctness of the AI. With further research, the AI Peer paradigm described here could open new possibilities for how we learn, adapt to, and grow with AI.
A Generative Approach to LLM Harmfulness Detection with Red Flag Tokens
Sophie Xhonneux
David Dobre
Mehrnaz Mofakhami
Leo Schwinn
Most safety training methods for large-language models (LLMs) based on fine-tuning rely on dramatically changing the output distribution of … (voir plus)the model when faced with a harmful request, shifting it from an unsafe answer to a refusal to respond. These methods inherently compromise model capabilities and might make auto-regressive models vulnerable to attacks that make likely an initial token of affirmative response. To avoid that, we propose to expand the model's vocabulary with a special token we call a *red flag token* (
Grokking Beyond the Euclidean Norm of Model Parameters
Tikeng Notsawo Pascal Junior
Grokking refers to a delayed generalization following overfitting when optimizing artificial neural networks with gradient-based methods. I… (voir plus)n this work, we demonstrate that grokking can be induced by regularization, either explicit or implicit. More precisely, we show that when there exists a model with a property
A Guide to Misinformation Detection Data and Evaluation
Camille Thibault
Jacob-Junqi Tian
Gabrielle Péloquin-Skulski
Taylor Lynn Curtis
Florence Laflamme
James Zhou
Yuxiang Guan
Kellin Pelrine
Misinformation is a complex societal issue, and mitigating solutions are difficult to create due to data deficiencies. To address this probl… (voir plus)em, we have curated the largest collection of (mis)information datasets in the literature, totaling 75. From these, we evaluated the quality of all of the 36 datasets that consist of statements or claims, as well as the 9 datasets that consists of data in purely paragraph form. We assess these datasets to identify those with solid foundations for empirical work and those with flaws that could result in misleading and non-generalizable results, such as insufficient label quality, spurious correlations. We further provide state-of-the-art baselines on all these datasets, but show that regardless of label quality, categorical labels may no longer give an accurate evaluation of detection model performance. We discuss alternatives to mitigate this problem. Overall, this guide aims to provide a roadmap for obtaining higher quality data and conducting more effective evaluations, ultimately improving research in misinformation detection. All datasets and other artifacts are available at [anonymized].
Landscape of Thoughts: Visualizing the Reasoning Process of Large Language Models
Zhanke Zhou
Xuan Li
Zhaocheng Zhu
Mikhail Galkin
Xiao Feng
Sanmi Koyejo
Bo Han
Numerous applications of large language models (LLMs) rely on their ability to perform step-by-step reasoning. However, the reasoning behavi… (voir plus)or of LLMs remains poorly understood, posing challenges to research, development, and safety. To address this gap, we introduce landscape of thoughts-the first visualization tool for users to inspect the reasoning paths of chain-of-thought and its derivatives on any multi-choice dataset. Specifically, we represent the states in a reasoning path as feature vectors that quantify their distances to all answer choices. These features are then visualized in two-dimensional plots using t-SNE. Qualitative analysis shows that the landscape of thoughts effectively distinguishes between strong and weak models, correct and incorrect answers, as well as different reasoning tasks. It also uncovers undesirable reasoning patterns, such as low consistency and high uncertainty. Additionally, users can adapt our tool to a neural model that predicts any property they observe. We showcase this advantage by adapting our tool to a lightweight verifier, which significantly improves reasoning by evaluating the correctness of reasoning paths.
Landscape of Thoughts: Visualizing the Reasoning Process of Large Language Models
Zhanke Zhou
Zhaocheng Zhu
Xuan Li
Mikhail Galkin
Xiao Feng
Sanmi Koyejo
Bo Han
Learning Decision Trees as Amortized Structure Inference
Mohammed Mahfoud
Ghait Boukachab
Michał Koziarski
Alex Hernandez-Garcia
Stefan Bauer
Nikolay Malkin
Learning to Defer for Causal Discovery with Imperfect Experts
Oscar Clivio
Divyat Mahajan
Perouz Taslakian
Sara Magliacane
Valentina Zantedeschi
Integrating expert knowledge, e.g. from large language models, into causal discovery algorithms can be challenging when the knowledge is not… (voir plus) guaranteed to be correct. Expert recommendations may contradict data-driven results, and their reliability can vary significantly depending on the domain or specific query. Existing methods based on soft constraints or inconsistencies in predicted causal relationships fail to account for these variations in expertise. To remedy this, we propose L2D-CD, a method for gauging the correctness of expert recommendations and optimally combining them with data-driven causal discovery results. By adapting learning-to-defer (L2D) algorithms for pairwise causal discovery (CD), we learn a deferral function that selects whether to rely on classical causal discovery methods using numerical data or expert recommendations based on textual meta-data. We evaluate L2D-CD on the canonical Tübingen pairs dataset and demonstrate its superior performance compared to both the causal discovery method and the expert used in isolation. Moreover, our approach identifies domains where the expert's performance is strong or weak. Finally, we outline a strategy for generalizing this approach to causal discovery on graphs with more than two variables, paving the way for further research in this area.
Learning to Defer for Causal Discovery with Imperfect Experts
Oscar Clivio
Divyat Mahajan
Perouz Taslakian
Sara Magliacane
Valentina Zantedeschi
Integrating expert knowledge, e.g. from large language models, into causal discovery algorithms can be challenging when the knowledge is not… (voir plus) guaranteed to be correct. Expert recommendations may contradict data-driven results, and their reliability can vary significantly depending on the domain or specific query. Existing methods based on soft constraints or inconsistencies in predicted causal relationships fail to account for these variations in expertise. To remedy this, we propose L2D-CD, a method for gauging the correctness of expert recommendations and optimally combining them with data-driven causal discovery results. By adapting learning-to-defer (L2D) algorithms for pairwise causal discovery (CD), we learn a deferral function that selects whether to rely on classical causal discovery methods using numerical data or expert recommendations based on textual meta-data. We evaluate L2D-CD on the canonical Tübingen pairs dataset and demonstrate its superior performance compared to both the causal discovery method and the expert used in isolation. Moreover, our approach identifies domains where the expert's performance is strong or weak. Finally, we outline a strategy for generalizing this approach to causal discovery on graphs with more than two variables, paving the way for further research in this area.
Mol-MoE: Training Preference-Guided Routers for Molecule Generation
Diego Calanzone
Pierluca D'Oro
Recent advances in language models have enabled framing molecule generation as sequence modeling. However, existing approaches often rely on… (voir plus) single-objective reinforcement learning, limiting their applicability to real-world drug design, where multiple competing properties must be optimized. Traditional multi-objective reinforcement learning (MORL) methods require costly retraining for each new objective combination, making rapid exploration of trade-offs impractical. To overcome these limitations, we introduce Mol-MoE, a mixture-of-experts (MoE) architecture that enables efficient test-time steering of molecule generation without retraining. Central to our approach is a preference-based router training objective that incentivizes the router to combine experts in a way that aligns with user-specified trade-offs. This provides improved flexibility in exploring the chemical property space at test time, facilitating rapid trade-off exploration. Benchmarking against state-of-the-art methods, we show that Mol-MoE achieves superior sample quality and steerability.
PREFERENCE OPTIMIZATION FOR CONCEPT BOTTLENECK MODELS
Emiliano Penaloza
Tianyue H. Zhang
Mateo Espinosa Zarlenga
Concept Bottleneck Models (CBMs) propose to enhance the trustworthiness of AI systems by constraining their decisions on a set of human-unde… (voir plus)rstandable concepts. However, CBMs typically assume that datasets contain accurate concept labels—an assumption often violated in practice, which we show can significantly degrade performance (by 25% in some cases). To address this, we introduce the Concept Preference Optimization (CPO) objective, a new loss function based on Direct Preference Optimization, which effectively mitigates the negative impact of concept mislabeling on CBM performance. We provide an analysis of some key properties of the CPO objective showing it directly optimizes for the concept’s posterior distribution, and contrast it against Binary Cross Entropy (BCE) where we show CPO is inherently less sensitive to concept noise. We empirically confirm our analysis finding that CPO consistently outperforms BCE in three real-world datasets with and without added label noise.