NeurIPS 2024PastInterpretability
Second NeurIPS Workshop on Attributing Model Behavior at Scale
ATTRIB 2024
- Submission deadline
- Oct 5, 2024, 12:00 UTCimported from OpenReview — check the website for extensions
- Submission portal
- OpenReview
- Notes
- Topics were auto-suggested and may be imprecise — edits welcome.
Accepted papers (72)
Fetched from OpenReview (v2) on 2026-06-10.
$\Delta$-Influence: Unlearning Poisons via Influence Functions
$\texttt{dattri}$: A Library for Efficient Data Attribution
A Comparative Study of Translation Bias and Accuracy in Multilingual Large Language Models for Cross-Language Claim Verification
A Versatile Influence Function for Data Attribution with Non-Decomposable Loss
Ablation is Not Enough to Emulate DPO: How Neuron Dynamics Drive Toxicity Reduction
Accumulated Local Effects for Link Prediction with Graph Neural Networks
Accumulating Data Avoids Model Collapse
Activation Monitoring: Advantages of Using Internal Representations for LLM Oversight
Adversarial Attacks on Data Attribution
Algorithmic Phase Transitions in Large Language Models: A Mechanistic Case Study of Arithmetic
Approximations to worst-case data dropping: unmasking failure modes
Attributing Statistics to Synthesis Quality in Correlation-Based Texture Models
BAKU: An Efficient Transformer for Multi-Task Policy Learning
Better Counterfactual Model Reasoning with Submodular Quadratic Component Models
Between the Bars: Gradient-based Jailbreaks are Bugs that induce Features
Bias Analysis for Unconditional Image Generative Models
Common Functional Decompositions Can Mis-attribute Differences in Outcomes Between Populations
Converging to a Lingua Franca: Evolution of Linguistic Regions and Semantics Alignment in Multilingual Large Language Models
Data Attribution for Multitask Learning
Decoding Dark Matter: Specialized Sparse Autoencoders for Interpreting Rare Concepts in LLMs
Detecting Origin Attribution for Text-to-Image Diffusion Models in RGB and Beyond
Efficient Ensembles Improve Training Data Attribution
Evaluating Sparse Autoencoders for Controlling Open-Ended Text Generation
Evaluating Sparse Autoencoders on Targeted Concept Removal Tasks
Evaluating Synthetic Activations composed of SAE Latents in GPT-2
Evolution of SAE Features Across Layers in LLMs
Feature Responsiveness Scores: Model-Agnostic Explanations for Agency
Final-Model-Only Data Attribution with a Unifying View of Gradient-Based Methods
From Loops to Oops: Fallback Behaviors of Language Models Under Uncertainty
Generalization v.s. Memorization: Tracing Language Models’ Capabilities Back to Pretraining Data
Generalized Group Data Attribution
GPT-2 Through the Lens of Vector Symbolic Architectures
GRADE: A Fine-grained Approach to Measure Sample Diversity in Text-to-Image Models
Hessian Sets: Uncovering Feature Interactions in Image Classification
How Many Van Goghs Does It Take to Van Gogh? Finding the Imitation Threshold
How much can we forget about Data Contamination?
In Search of Forgotten Domain Generalization
Inconsistencies In Consistency Models: Better ODE Solving Does Not Imply Better Samples
Inductive Linguistic Reasoning with Large Language Models
Influence Functions for Scalable Data Attribution in Diffusion Models
Influence-based Attributions can be Manipulated
Investigating Language Model Dynamics using Meta-Tokens
Investigating Sensitive Directions in GPT-2: An Improved Baseline and Comparative Analysis of SAEs
Just Select Twice: Leveraging Low Quality Data to Improve Data Selection
Latent Concept-based Explanation of NLP Models
Loss-to-Loss Prediction: Language model scaling laws across datasets
Most Influential Subset Selection: Challenges, Promises, and Beyond
On Linear Representations and Pretraining Data Frequency in Language Models
Peter Parker or Spiderman? Disambiguating Multiple Class Labels
Pruning-based Data Selection and Network Fusion for Efficient Deep Learning
Quanda: An Interpretability Toolkit for Training Data Attribution Evaluation and Beyond
Quantifying Positional Biases in Text Embedding Models
ReLU's Revival: On the Entropic Overload in Normalization-Free Large Language Models
SCIURus: Shared Circuits for Interpretable Uncertainty Representations in Language Models
Secret Seeds in Text-to-Image Diffusion Models
Small-to-Large Generalization: Training Data Influences Models Consistently Across Scale
The Association Between Training Data and Text-to-Image Generation Capabilities
Toward Optimal Search and Retrieval for RAG
Towards a Mechanistic Explanation of Diffusion Model Generalization
Towards Data-Centric RLHF: Simple Metrics for Preference Dataset Comparison
Training on the Test Task Confounds Evaluation and Emergence
U-shaped and Inverted-U Scaling behind Emergent Abilities of Large Language Models
Understanding Compute-Parameter Trade-offs in Sparse Mixture-of-Expert Language Models
Understanding the Sources of Performance in Deep Drug Response Models
Unintentional Unalignment: Likelihood Displacement in Direct Preference Optimization
Visual Reasoning Requires Rethinking Vision-Language Beyond Scaling
Weak-to-Strong Confidence Prediction
Weak-to-Strong In-Context Optimization of Language Model Reasoning
What do Learning Dynamics Reveal about Generalization in LLM Reasoning?
What's In My Big Data?
When Attention Sink Emerges in Language Models: An Empirical View
You can remove GPT2's LayerNorm by fine-tuning