Papers
12094 papers
ICLR2025
Can We Talk Models Into Seeing the World Differently?
Summary pending...
vision language modelsvision biasesshape/texture bias
ICLR2025
When is Task Vector Provably Effective for Model Editing? A Generalization Analysis of Nonlinear Transformers
Summary pending...
Task arithmeticgeneralizationnonlinear Transformers
ICLR2025
Systems with Switching Causal Relations: A Meta-Causal Perspective
Summary pending...
Meta-CausalityMeta-Causal ReasoningAgent Behavior
ICLR2025
ACES: Automatic Cohort Extraction System for Event-Stream Datasets
Summary pending...
Automatic Task SpecificationCohort ExtractionElectronic Health Records
ICLR2025
Inference-Aware Fine-Tuning for Best-of-N Sampling in Large Language Models
Summary pending...
Best-of-N samplingReinforcement LearningLanguage models
ICLR2025
DeFT: Decoding with Flash Tree-attention for Efficient Tree-structured LLM Inference
Summary pending...
LLM inferenceattentionmemory-efficiency
ICLR2025
Intermediate Layer Classifiers for OOD generalization
Summary pending...
transfer learningintermediate layerslearning dynamics
ICLR2025
Robust System Identification: Finite-sample Guarantees and Connection to Regularization
Summary pending...
dynamical systemtime seriessystem identification
ICLR2025
Robust Feature Learning for Multi-Index Models in High Dimensions
Summary pending...
feature learningadversarial robustnessneural networks
ICLR2025
Flow Matching with Gaussian Process Priors for Probabilistic Time Series Forecasting
Summary pending...
flow matchingtime series forecastinggenerative modeling
ICLR2025
LoLCATs: On Low-Rank Linearizing of Large Language Models
Summary pending...
Linear AttentionLinearizing TransformersLow-rank Adaptation
ICLR2025
Protein Language Model Fitness is a Matter of Preference
Summary pending...
protein language modelszero-shot fitness prediction
ICLR2025
Self-Play Preference Optimization for Language Model Alignment
Summary pending...
self playpreference optimizationlarge language model
ICLR2025
Improving Pretraining Data Using Perplexity Correlations
Summary pending...
pretraining data; data selection; natural language processing; statistics; large language models
ICLR2025
Single-agent Poisoning Attacks Suffice to Ruin Multi-Agent Learning
Summary pending...
Multi-agent learningreward poisoning attackNash equilibrium
ICLR2025
Nonlinear Sequence Embedding by Monotone Variational Inequality
Summary pending...
Monotone Variational InequalityConvex OptimizationSequence Data
ICLR2025
Broadening Target Distributions for Accelerated Diffusion Models via a Novel Analysis Approach
Summary pending...
generative modelsdenoising diffusion probabilistic model (DDPM)convergence analysis
ICLR2025
On the self-verification limitations of large language models on reasoning and planning tasks
Summary pending...
Large Language ModelsReasoningPlanning
ICLR2025
Round and Round We Go! What makes Rotary Positional Encodings useful?
Summary pending...
Large Language ModelsTransformersPositional Encodings
ICLR2025
BenTo: Benchmark Reduction with In-Context Transferability
Summary pending...
transfer learninglanguage modelbenchmark reduction