Publications
Peer-reviewed papers, preprints and theses from the lab. Lab members are highlighted.
2026
-
ICLR 2026
Implicit Regularization of SGD Reduces Shortcut Learning
Shows how the implicit regularization of SGD, controlled by learning rate and batch size, suppresses reliance on spurious features while preserving accuracy.
International Conference on Learning Representations (ICLR), 2026
-
TMLR 2026
Bridging Reasoning to Learning: Unmasking Illusions Using Complexity Out-of-Distribution Generalization
Frames reasoning as generalization to problems whose solutions require greater representational or computational complexity than the training examples.
Transactions on Machine Learning Research (TMLR), 2026
-
TMLR 2026
CARINOX: Inference-Time Scaling with Category-Aware Reward-Based Initial Noise Optimization and Exploration
Improves compositional text-to-image generation by optimizing and exploring initial noise with category-aware rewards at inference time.
Transactions on Machine Learning Research (TMLR), 2026
-
ICASSP 2026
No Concept Left Behind: Test-Time Optimization for Compositional Text-to-Image Generation
Optimizes generation at test time to preserve every requested concept and improve compositional prompt fidelity.
IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2026
-
EACL Findings 2026
MEENA (PersianMMMU): Multimodal-Multilingual Educational Exams for N-Level Assessment
Introduces a large bilingual Persian–English benchmark for evaluating vision-language models across educational, scientific, and cultural reasoning tasks.
Findings of the Association for Computational Linguistics: EACL, 2026
-
ICML 2026
HaloProbe: Bayesian Detection and Mitigation of Object Hallucinations in Vision-Language Models
Uses Bayesian probing to detect object hallucinations and guide mitigation in vision-language model responses.
International Conference on Machine Learning (ICML), 2026
-
arXiv 2026
Hidden Meanings in Plain Sight: RebusBench for Evaluating Cognitive Visual Reasoning
Introduces a rebus-puzzle benchmark that tests whether multimodal models can integrate visual form, language, and abstract concepts.
arXiv preprint, 2026
-
ACL Findings 2026
Mechanistic Interpretability of Large-Scale Counting in LLMs through a System-2 Strategy
Reveals how a System-2 strategy decomposes large counting problems, transfers partial counts through dedicated attention heads, and aggregates them reliably.
Findings of the Association for Computational Linguistics: ACL, 2026
-
arXiv 2026
Causal Attribution via Activation Patching
Develops activation-patching methods for tracing causal contributions inside neural networks rather than relying on correlational explanations.
arXiv preprint, 2026
-
TMLR 2026
Large Language Models for Scientific Idea Generation: A Creativity-Centered Survey
Surveys the use of large language models for scientific idea generation through a creativity-centered framework and evaluation perspective.
Transactions on Machine Learning Research (TMLR), 2026
-
arXiv 2026
Lying to Win: Assessing LLM Deception through Human-AI Games and Parallel-World Probing
Evaluates strategic deception through interactive games and probes whether models represent truthful and deceptive alternatives internally.
arXiv preprint, 2026
-
arXiv 2026
The Judge Who Never Admits: Hidden Shortcuts in LLM-Based Evaluation
Shows that language-model judges can exploit hidden shortcuts while producing explanations that fail to acknowledge those cues.
arXiv preprint, 2026
-
ICLR 2026
SUSD: Structured Unsupervised Skill Discovery through State Factorization
Factorizes environment states and skill variables to learn richer, more diverse, and compositionally reusable behaviors without external rewards.
International Conference on Learning Representations (ICLR), 2026
-
ICLR 2026
Efficient Adversarial Attacks on High-Dimensional Offline Bandits
Demonstrates how small targeted perturbations to reward models can hijack offline bandits and studies a practical partial defense.
International Conference on Learning Representations (ICLR), 2026
2025
-
NeurIPS 2025
Spurious-Aware Prototype Refinement for Reliable Out-of-Distribution Detection
Refines class prototypes with explicit awareness of spurious features to make out-of-distribution detection more reliable under shortcut-driven shifts.
Advances in Neural Information Processing Systems (NeurIPS), 2025
-
Machine Learning 2025
Inductive Biases for Zero-Shot Systematic Generalization in Language-Informed Reinforcement Learning
Studies compositional inductive biases that help language-informed reinforcement learning agents generalize systematically to unseen task combinations.
Machine Learning (Springer), 2025
-
TMLR 2025
CAREL: Instruction-Guided Reinforcement Learning with Cross-Modal Auxiliary Objectives
Uses cross-modal auxiliary objectives and instruction tracking to improve sample efficiency and systematic generalization in language-guided reinforcement learning.
Transactions on Machine Learning Research (TMLR), 2025
-
2025 Position paper
Position: System-2 AI Is About Complexity Out of Distribution
Argues that System-2 behavior is best understood as generalization beyond the computational complexity represented in training examples.
Position paper, 2025
-
ICML HiLD 2025
The Silent Helper: How Implicit Regularization Enhances Group Robustness
Studies how the implicit bias of optimization can improve worst-group performance even when group annotations are unavailable.
ICML Workshop on High-dimensional Learning Dynamics (HiLD), 2025
-
ICLR SCSL 2025
Exploiting What Trained Models Learn for Making Them Robust to Spurious Correlations without Group Annotations
Uses information already encoded by trained models to reduce shortcut reliance without requiring group labels during training.
ICLR Workshop on Spurious Correlation and Shortcut Learning (SCSL), 2025
-
TMLR 2025
Attention Overlap Is Responsible for the Entity Missing Problem in Text-to-Image Diffusion Models!
Connects missing entities in generated images to overlapping attention maps and develops an inference-time strategy for better object coverage.
Transactions on Machine Learning Research (TMLR), 2025
-
TMLR 2025
Why Settle for Mid: A Probabilistic Viewpoint to Spatial Relationship Alignment in Text-to-Image Models
Takes a probabilistic view of spatial relationships to improve how text-to-image models place and align objects described in prompts.
Transactions on Machine Learning Research (TMLR), 2025
-
arXiv 2025
Infinity and Beyond: Compositional Alignment in VAR and Diffusion T2I Models
Compares compositional alignment in visual autoregressive and diffusion models to expose their distinct strengths and failure modes.
arXiv preprint, 2025
-
arXiv 2025
Evaluating the Evaluators: Metrics for Compositional Text-to-Image Generation
Systematically evaluates whether popular metrics reliably measure object, attribute, and relationship alignment in generated images.
arXiv preprint, 2025
-
arXiv 2025
T2I-FineEval: Fine-Grained Compositional Metric for Text-to-Image Evaluation
Introduces a fine-grained metric for measuring whether generated images satisfy the individual components of complex prompts.
arXiv preprint, 2025
-
arXiv 2025
Fine-Grained Alignment and Noise Refinement for Compositional Text-to-Image Generation
Refines alignment signals and initial noise to improve compositional correctness without retraining the text-to-image model.
arXiv preprint, 2025
-
arXiv 2025
Diffusion Beats Autoregressive: An Evaluation of Compositional Generation in Text-to-Image Models
Benchmarks diffusion and autoregressive image generators and analyzes their relative ability to follow compositional prompts.
arXiv preprint, 2025
-
NeurIPS 2025
Visual Structures Help Visual Reasoning: Addressing the Binding Problem in LVLMs
Introduces VISER, which adds lightweight spatial structure and sequential-scanning prompts to improve counting, visual search, and spatial reasoning.
Advances in Neural Information Processing Systems (NeurIPS), 2025
-
CVPR 2025
CLIP Under the Microscope: A Fine-Grained Analysis of Multi-Object Representation
Provides a controlled analysis of how CLIP represents multiple objects and reveals limitations that aggregate benchmarks can conceal.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
-
arXiv 2025
Uncovering Grounding IDs: How External Cues Shape Multimodal Binding
Studies how external cues create grounding identifiers that help multimodal models bind objects to their properties and locations.
arXiv preprint, 2025
-
arXiv 2025
Analyzing CLIP's Performance Limitations in Multi-Object Scenarios: A Controlled High-Resolution Study
Uses controlled high-resolution data to isolate CLIP's limitations when images contain multiple objects and attributes.
arXiv preprint, 2025
-
CVPR 2025
LibraGrad: Balancing Gradient Flow for Universally Better Vision Transformer Attributions
Balances gradient flow to produce more stable, faithful, and broadly effective attribution maps for vision transformer architectures.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
-
arXiv 2025
Understanding Counting Mechanisms in Large Language and Vision-Language Models
Compares internal counting strategies across language and vision-language models to reveal shared mechanisms and modality-specific failures.
arXiv preprint, 2025
-
ACL 2025
CER: Confidence Enhanced Reasoning in LLMs
Uses model confidence to improve reasoning reliability and help language models make better decisions during multi-step inference.
Annual Meeting of the Association for Computational Linguistics (ACL), 2025
-
arXiv 2025
The Illusion of Procedural Reasoning: Measuring Long-Horizon FSM Execution in LLMs
Tests long-horizon finite-state-machine execution to separate genuine procedural reasoning from shallow pattern matching.
arXiv preprint, 2025
-
arXiv 2025
Unspoken Hints: Accuracy Without Acknowledgement in LLM Reasoning
Examines cases where implicit hints improve answers even though model explanations do not acknowledge using them.
arXiv preprint, 2025
-
arXiv 2025
The Silent Judge: Unacknowledged Shortcut Bias in LLM-as-a-Judge
Demonstrates that LLM evaluators can rely on shortcut signals while presenting apparently independent justifications for their scores.
arXiv preprint, 2025
-
IEEE Access 2025
Test-Time Confidence-Aware Query Labeling Improves 3D Few-Shot Segmentation
Uses confidence-aware query labeling at test time to improve the accuracy and adaptability of three-dimensional few-shot segmentation.
IEEE Access, 2025
-
MedIA 2025
Self-Supervised 3D Medical Image Segmentation by Flow-Guided Mask Propagation Learning
Learns three-dimensional medical image segmentation without dense labels by propagating masks through flow-guided self-supervision.
Medical Image Analysis, 2025
-
arXiv 2025
VQEL: Enabling Self-Developed Symbolic Language in Agents through Vector Quantization in Emergent Language Games
Uses vector quantization to help interacting agents develop a discrete symbolic language in emergent communication games.
arXiv preprint, 2025
No publications match your filters.