Research
Papers and datasets worth knowing, ranked by significance and community attention.
U-Space: Uncovering When and Why Uncertainty Arises in Language Models
Large language models are informing decisions with ever-higher stakes.
FearCaut-Qwen: Affective Steering in a Vision-Language Model Shifts the Decision Criterion for Hazard Assessment
Vision-language models (VLMs) show great potential for damage assessment after a disaster, but a recurring deficiency is that they are reluctant to declare a hazard; that is, recall is low even when overall accuracy appears adequate.
DSReg: Provably Recovering Individual World Latents without Reconstruction
Methods that recover individual latent variables of the world, from nonlinear ICA to dictionary learning and causal representation learning, anchor the latents to observations through reconstruction, auxiliary supervision, or distributional asymmetries such as non-Gaussianity.
MotherTree: Meta-learning on synthetic data improves decision tree training
We introduce MotherTree, a tabular transformer that meta-learns decision tree induction: given a training set for a new task, it outputs a hard, axis-aligned decision tree, equivalent in form to classically trained trees, in a single forward pass.
REFLECT: Role-Aware Detection of Sentiment, Emotion, Politeness, and Formality
Detecting communicative meaning in dialogue requires modeling not only lexical content but also the participant-role structures through which sentiment, emotion, politeness, and formality (SEPF) are expressed.
Transformer-based In-Context Score Operator Learning for Diffusion Models
We introduce an in-context solution operator learning framework for diffusion models, in which each target distribution is treated as a task drawn from a meta-distribution and is specified at inference time by an empirical context set.
Exploring Attention Sinks in Text-to-Image Transformers from Diffusion and Autoregressive Models
To fill this gap and better understand how text-to-image Transformers allocate and route attention, we explore this phenomenon of text-to-image transformers across bidirectional diffusion models and causal autoregressive models.
Evidential Multimodal Transformers with Dempster-Shafer Fusion for Emotion Recognition
We present an Evidential Multimodal Transformer with Dempster-Shafer Fusion (EMT-DS) that addresses this problem through formal uncertainty quantification.
A Stevens's Power Law Check-up of GPT-5.5's Implicit Reading of Visual Encoding
We adapt Stevens's power law to measure the implicit ability of AI models to read visualizations, which can reveal the built-in perceptual mechanisms of algorithmic models.
Marformer: A Transformer for Predicting Missing Data Distributions
We present the Marformer, a Transformer trained to directly predict conditional marginals given any set of observed values.
Optimization Encoders: Rethinking Second-Order Meta-Learning for Neural Fields
Guided by this view, we introduce Attentive Latent Fields (MetaLF), an equivariant transformer-based neural field that contextualizes a latent pointcloud through self-attention.
Why DiT Models Underperform as Representation Learners without Long Skip Connections
Diffusion feature, i.e., intermediate activations extracted from diffusion model backbones, has emerged as a promising approach for unifying generation and vision understanding (discrimination).
asdex: Automatic Sparse Differentiation in JAX
Automatic sparse differentiation (ASD) exploits this structure in four steps: detection of the input-agnostic sparsity pattern, coloring of a graph to group columns or rows that can share an AD pass, compressed differentiation to compute a compressed derivative matrix with one AD pass per color, and finally decompression into the original sparsity pattern.
Dissecting Representation Structure in Vision Transformers: A Rigorous Architectural Study
Representation structure is crucial for understanding Vision Transformer (ViT) architectures and their generalization behavior.
RouterInterp: Understanding Superposed Specialisation in Mixture of Experts Routing
Leveraging the SSH, we introduce RouterInterp, a method for interpreting expert routing that identifies Sparse Autoencoder features most predictive of routing decisions and produces unified natural language explanations.
QCLR: A Plug-in Robustification Module via Quantization for Contrastive Learning in Self-Supervised Skeleton-Based Human Action Recognition
In this work, we propose a plug-in robustification module via quantization for contrastive learning in self-supervised skeleton-based human action recognition (QCLR), which enhances the robustness of representations against adversarial attacks while preserving their representational capability.
How Do Transformers Learn to Represent Symmetries?
Training Transformer-based architectures with finite data augmentation has become an increasingly popular approach in geometric machine learning.
Fully Interpretable Minimal Transformers: From Geometry to Algorithm
We present a framework for building and interpreting minimal transformer models.
READ: Reading the Selective Scan for Faithful Video Attribution
Selective state-space models are an efficient approach to long-sequence video modeling.
Leaner Transformers Can Easily Learn to Cluster
Recent work shows that transformers can exactly perform Lloyd's algorithm for $k$-means clustering with $n$ points in $d$ dimensions with an embedding size $d{\textsf{emb}} = d+k$ (thus, requiring attention projection matrices of size $(d+k)^2$).
Defence against the dark arts of distillation: Unsupervised discovery of subliminal learning
We remove this requirement and prove that a subliminally transferred trait can be recovered from the parameter updates alone, without supervision, by exploiting a single asymmetry: the trait is expressed consistently across diverse datasets generated by the teacher, whereas task-specific and shared capabilities vary.
Temporal transformer CAN encoder with federated lightweight heads for anomaly detection
Motivated by this gap, we present a privacy-preserving framework for anomaly detection in in-vehicle networks, based on a Temporal Transformer CAN Encoder with Federated Lightweight Heads, to better capture these irregularities.
Static Gradient Attribution Underperforms a Density-Matched Random Mask on Loss-Based Forgetting Within LoRA B-Matrix
Parameter-efficient continual learning updates a restricted subset of adapter parameters to incorporate sequential tasks while limiting catastrophic forgetting.
Memorization and Malign Generalization in Conditional Diffusion Models with Random Features
Conditional diffusion models generate diverse, novel, and high-quality samples under prescribed conditions.
Attributing HOW, Not Just WHICH: Counterfactual Response Trajectories for Diffusion Models
Diffusion models have achieved remarkable success in image generation, yet tracing their outputs to individual training examples remains challenging.
CorrFlow: Exact and Permutation-Equivariant Flow Matching on the Correlation Elliptope
Correlation matrices combine nonlinear positive-definiteness constraints with a simultaneous row-column permutation symmetry.
PlacePulse: Learning Transferable Place Representations from Structural Geographic Data
We introduce PlacePulse, a structural geospatial foundation model that compresses 572 attributes, expanded to 2,288 inputs through H3 neighborhood aggregates, into 256-dimensional unit-norm embeddings for 1,055,537 H3 resolution-7 cells in the United States.
Bridging Human and Model Attention: Explainability Analysis of CNN, Mamba, and ViT Architectures with Gaze-Based Validation
We conduct a systematic explainability evaluation across Convolutional Neural Networks (CNNs), Vision Transformers (ViTs), and Vision Mamba (ViM).
Rubric-CEPR: Self-Evolving Image Editing via Reward-Verified Self-Distillation
To this end, we propose a self-evolving framework, named Rubric-CEPR, that verifies the editor's own samples with its internal representations through a rubric-augmented Contrastive Edit-Preservation Reward (CEPR).
An Interpretable Approach to PDE Solution Discovery via Structural Experience Distillation
PDE solution discovery aims to identify explicit symbolic expressions for unknown physical fields from observations under known physical constraints.
BEACON-SP: Ontology-Grounded GraphRAG Framework for Clinical Suicide Risk Assessment
We present BEACON-SP, an ontology-grounded Graph Retrieval-Augmented Generation (GraphRAG) framework for clinician-facing decision support in behavioral health settings such as suicide prevention, where effective assessment requires integrating heterogeneous clinical, behavioral, social, and temporal evidence.
Chaos in the Text: Revealing the Modality Preference in Mixed-Modality Retrievers
Dense retrievers have made significant progress on text and image corpora, but whether these capabilities extend reliably to mixed corpora containing text, image, and fused text-image documents remains unclear.
Adversarially Trained Linear Transformers Are Optimal Robust In-Context Learners for Gaussian Mixtures
Specifically, we show that, for a family of Gaussian-mixture classification tasks, a sufficiently deep linear transformer adversarially trained across tasks can asymptotically attain the robust Bayes error on previously unseen tasks through in-context learning from clean demonstrations.
Is In-Domain Training Enough for Fine-Grained Industrial Anomaly Understanding?
A single multimodal large language model (MLLM) struggles to excel simultaneously at detection, localization, description, and reasoning in multimodal industrial anomaly understanding (MM-IAU).
Stability of Measure-to-Measure Transformers on Sub-Gaussian Data
We show that transformers map sub-Gaussian inputs to sub-Gaussian outputs; this ensures that taking arbitrary-length compositions of the softmax operator is well-defined.
Rethinking Contrastive Loss in CLIP Post-training: A Complementary Framework with Frozen Text Encoder
Building on this finding, we propose ComCLIP, a lightweight single-epoch post-training recipe that freezes CLIP's text encoder---so the refined vision encoder is a drop-in replacement with unchanged architecture and inference cost---and trains the vision encoder with a properly-tempered contrastive loss, an MSE anchoring loss against the original CLIP, and a relational distillation loss from DINOv2.
AutoAdapt: Automatic Domain Discovery Enables Low-Cost Extensibility
We present AutoAdapt, a modular framework that incorporates new domains and data via targeted single-adapter training without modifying other adapters.
CARE: Constrained Attention Refinement for Fine-Grained Visual Classification via Teacher-Student Distillation
To address this problem, we propose CARE, a constrained attention refinement framework for interpretable fine-grained recognition via teacher-student distillation.
When Citations Mislead? A Claim-Level Benchmark for Legal Hallucination Detection
We introduce PARCEL, a benchmark for checking whether a legal claim is supported by the underlying authority.
MiniVer-V: Identifying Minimal Sufficient Evidence for Short Video Verification
We introduce MiniVer-V, a benchmark of 195 short videos with three-way verdict annotations (supported, refuted, insufficient) and 5,510 multimodal evidence units spanning visual keyframes, speech transcripts, and web-retrieved external sources.
Anchor-driven Multi-modal Multi-scale Expert Selection for Survival Prediction
To address these limitations, we propose an Anchor-driven Multi-modal Multi-scale Expert Selection (AM$^2$ES) framework for survival prediction.
Not All Dimensions Are Equal: Outlier-Aware Attention Linearization with Grouped Quadratic Kernels
Motivated by this heterogeneity, we propose Outlier-Aware Attention Linearization (OAL), a dimension-aware method for constructing quadratic attention approximations.
Decoding Register Tokens Preserves Global Scene Details
Pretrained vision transformers learn structured image representations spanning spatial patch tokens and non-spatial tokens.
Quantize by Drift: Label-Free Mixed-Precision Post-Training Quantization for Text Embedders
Mixed-precision post-training quantization needs a per-module sensitivity signal; for a text embedder the obvious one -- the retrieval quality a module costs when quantized -- needs relevance labels that deployments rarely have.
Detecting a Shift Is Not Enough: Exact Minimax Limits of Linear Representation Repair
We cast its removal as a statistical decision problem: from noisy differences between paired calibration measurements in $\mathbb{R}^d$, learn one linear map, applied to both sources under a hard distortion budget, that leaves as little of the shift as possible on fresh data.
CLUE: A Contrastive Learning Framework for Unsupervised Event Type Induction and Annotation
To alleviate the need for manual event ontology creation, we propose the CLUE framework for event type induction, naming, and annotation.
Patient, Place, Prior (P$^3$): What Counts as Personalization in Medical World Models?
We introduce Patient, Place, Prior (P$^3$), an audit asking whether a forecast benefits from the patient's longitudinal imaging history (Patient), benefits from patient-matched externally supplied spatial support (Place), and gains predictive value beyond a population-average prediction under matched support and context (Prior).
Multi-Stream Attention Residuals
Inspired by the success of multi-head attention along the sequence axis, we introduce Multi-Stream Attention Residuals (MSAR), an extension of AttnRes that generalizes this mechanism to multi-head depth attention, yielding substantial performance gains with minimal additional overhead.