Entity · paper

Influcoder: Distilling Decoders' Gradient Influence Rankings into an Encoder for Data Attribution

paperactiveinflucoder-distilling-decoders-gradient-influence-rankings-into-an-encoder-for-data-attribution-bee6034c·1 events·first seen Jun 12, 2026

Aliases: Influcoder: Distilling Decoders' Gradient Influence Rankings into an Encoder for Data Attribution

Co-occurring entities

Influcoder

More like this (12)

Graph Neural Network Encoder Sparse Autoencoders Encode Both Concepts and Functions: The Downstream Geometry of Feature Effects Integrated Gradients Gradient-Guided Reward Optimization Leveraging unlabelled data for generalizable neural population decoding Influcoder Feature Auto-Encoder Requential Coding: Pushing the Limits of Model Compression with Self-Generated Training Data Visual Attribution Distillation Beyond Independent Labels: Schwartz-Geometry Decoding for Human Value Detection Grading the Grader: Lessons from Evaluating an Agentic Data Analysis System Unstable Features, Reproducible Subspaces: Understanding Seed Dependence in Sparse Autoencoders

Recent events (1)

4arXiv · cs.CL·Jun 12, 2026·source ↗

Influcoder: Distilling gradient influence rankings into an encoder for scalable data attribution

Influcoder is a proposed method for scalable data attribution in LLM training, distilling decoder-based gradient influence rankings into a compact encoder representation. The approach targets the practical bottleneck of influence function methods — their high computational cost and storage requirements — making them viable for large-scale dataset curation. The work is relevant to training data quality filtering and identifying sources of undesirable model behavior such as toxicity.

Training Infrastructure Evaluation and Benchmarking Influcoder Influcoder: Distilling Decoders' Gradient Influence Rankings into an Encoder for Data Attribution