Image de couverture de livre

The Mirrored Influence Hypothesis: Efficient Data Influence Estimation by Harnessing Forward Passes

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Ko, Myeongseob, Kang, Feiyang, Shi, Weiyan, Jin, Ming, Yu, Zhou, Jia, Ruoxi
Format:	Preprint
Publié:	2024
Sujets:	Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Characterizing Model-Native Skills
par: Kang, Feiyang, et autres
Publié: (2026)

The Signal is in the Steps: Local Scoring for Reasoning Data Selection
par: Just, Hoang Anh, et autres
Publié: (2025)

Probing Knowledge Holes in Unlearned LLMs
par: Ko, Myeongseob, et autres
Publié: (2025)

Injecting Measurement Information Yields a Fast and Noise-Robust Diffusion-Based Inverse Problem Solver
par: Patsenker, Jonathan, et autres
Publié: (2025)

Boosting Alignment for Post-Unlearning Text-to-Image Generative Models
par: Ko, Myeongseob, et autres
Publié: (2024)

Capturing the Temporal Dependence of Training Data Influence
par: Wang, Jiachen T., et autres
Publié: (2024)

AdaDeDup: Adaptive Hybrid Data Pruning for Efficient Large-Scale Object Detection Training
par: Kang, Feiyang, et autres
Publié: (2025)

Accumulative SGD Influence Estimation for Data Attribution
par: Shi, Yunxiao, et autres
Publié: (2025)

AutoScale: Scale-Aware Data Mixing for Pre-Training LLMs
par: Kang, Feiyang, et autres
Publié: (2024)

DCFold: Efficient Protein Structure Generation with Single Forward Pass
par: Zhang, Zhe, et autres
Publié: (2026)

Data-Centric Human Preference with Rationales for Direct Preference Alignment
par: Just, Hoang Anh, et autres
Publié: (2024)

A Sustainable AI Economy Needs Data Deals That Work for Generators
par: Jia, Ruoxi, et autres
Publié: (2026)

Efficient Data Shapley for Weighted Nearest Neighbor Algorithms
par: Wang, Jiachen T., et autres
Publié: (2024)

f-INE: A Hypothesis Testing Framework for Estimating Influence under Training Randomness
par: Panda, Subhodip, et autres
Publié: (2025)

Retracing the Past: LLMs Emit Training Data When They Get Lost
par: Ko, Myeongseob, et autres
Publié: (2025)

DataInf: Efficiently Estimating Data Influence in LoRA-tuned LLMs and Diffusion Models
par: Kwon, Yongchan, et autres
Publié: (2023)

The Convergence Gap: Instruction-Tuned Language Models Stabilize Later in the Forward Pass
par: Zhou, Yifan
Publié: (2026)

HyperINF: Unleashing the HyperPower of the Schulz's Method for Data Influence Estimation
par: Zhou, Xinyu, et autres
Publié: (2024)

Training Data Influence Analysis and Estimation: A Survey
par: Hammoudeh, Zayd, et autres
Publié: (2022)

The Approximate Fisher Influence Function: Faster Estimation of Data Influence in Statistical Models
par: Lev, Omri, et autres
Publié: (2024)

Memory-Induced Tool-Drift in LLM Agents
par: Dabas, Mahavir, et autres
Publié: (2026)

Influence Dynamics and Stagewise Data Attribution
par: Lee, Jin Hwa, et autres
Publié: (2025)

Test-Time Model Adaptation with Only Forward Passes
par: Niu, Shuaicheng, et autres
Publié: (2024)

MATES: Model-Aware Data Selection for Efficient Pretraining with Data Influence Models
par: Yu, Zichun, et autres
Publié: (2024)

Influence Strength Estimation in Hyperbolic Space for Social Influence Maximization
par: Qiao, Hongliang, et autres
Publié: (2025)

Get more for less: Principled Data Selection for Warming Up Fine-Tuning in LLMs
par: Kang, Feiyang, et autres
Publié: (2024)

Automated Efficient Estimation using Monte Carlo Efficient Influence Functions
par: Agrawal, Raj, et autres
Publié: (2024)

Can We Trust the Performance Evaluation of Uncertainty Estimation Methods in Text Summarization?
par: He, Jianfeng, et autres
Publié: (2024)

Influence Functions for Efficient Data Selection in Reasoning
par: Humane, Prateek, et autres
Publié: (2025)

Layer-Aware Influence for Online Data Valuation Estimation
par: Yang, Ziao, et autres
Publié: (2025)

Data-Efficient RLVR via Off-Policy Influence Guidance
par: Zhu, Erle, et autres
Publié: (2025)

Data Attribution for Diffusion Models: Timestep-induced Bias in Influence Estimation
par: Xie, Tong, et autres
Publié: (2024)

Quagmires in SFT-RL Post-Training: When High SFT Scores Mislead and What to Use Instead
par: Kang, Feiyang, et autres
Publié: (2025)

Fine-Tuning Language Models with Just Forward Passes
par: Malladi, Sadhika, et autres
Publié: (2023)

Efficient Data Selection at Scale via Influence Distillation
par: Nikdan, Mahdi, et autres
Publié: (2025)

Just Enough Shifts: Mitigating Over-Refusal in Aligned Language Models with Targeted Representation Fine-Tuning
par: Dabas, Mahavir, et autres
Publié: (2025)

Data Descriptions from Large Language Models with Influence Estimation
par: Kim, Chaeri, et autres
Publié: (2025)

Provably Efficient Online RLHF with One-Pass Reward Modeling
par: Li, Long-Fei, et autres
Publié: (2025)

Which Experiences Are Influential for RL Agents? Efficiently Estimating The Influence of Experiences
par: Hiraoka, Takuya, et autres
Publié: (2024)

CHIPS: Efficient CLIP Adaptation via Curvature-aware Hybrid Influence-based Data Selection
par: Zhuang, Xinlin, et autres
Publié: (2025)