paper-with-me

홈 › Papers

Rethinking Test-Time Training: Tilting The Latent Distribution For Few-Shot Source-Free Adaptation

2026-02-02 · Tahir Qasim Syed, Behraj Khan arxiv

Often, constraints arise in deployment settings where even lightweight parameter updates e.g. parameter-efficient fine-tuning could induce model shift or tuning instability. We study test-time adaptation of foundation models for few-shot classification under a completely frozen-model regime, where additionally, no upstream data are accessible. We propose arguably the first training-free inference method that adapts predictions to the new task by performing a change of measure over the latent embedding distribution induced by the encoder. Using task-similarity scores derived from a small labeled support set, exponential tilting reweights latent distributions in a KL-optimal manner without modifying model parameters. Empirically, the method consistently competes with parameter-update-based methods across multiple benchmarks and shot regimes, while operating under strictly and universally stronger constraints. These results demonstrate the viability of inference-level distributional correction for test-time adaptation even with a fully-frozen model pipeline.

📄 PDF Abstract BibTeX arXiv:2602.02633

Code (0)

등록된 구현이 없습니다.

Tasks

parameter-efficient fine-tuningTest-time Adaptation

Similar Papers 제목 키워드 기반

Inference-Time Rethinking with Latent Thought Vectors for Math Reasoning

2026-02-06 · Deqian Kong, Minglu Zhao, Aoyang Qin, Bo Pang 외 arxiv

Standard chain-of-thought reasoning generates a solution in a single forward pass, committing irrevocably to each token and lacking a mechanism to recover from early errors. We introduce Inference-Time Rethinking, a gene…

Mathematical Reasoning

Reward Weighted Classifier-Free Guidance as Policy Improvement in Autoregressive Models

2026-04-16 · Alexander Peysakhovich, William Berman arxiv

Consider an auto-regressive model that produces outputs x (e.g., answers to questions, molecules) each of which can be summarized by an attribute vector y (e.g., helpfulness vs. harmlessness, or bio-availability vs. lipo…

Reinforcement Learning

Exponential Tilting of Generative Models: Improving Sample Quality by Training and Sampling from Latent Energy

2020-06-15 · Zhisheng Xiao, Qing Yan, Yali Amit

In this paper, we present a general method that can improve the sample quality of pre-trained likelihood based generative models. Our method constructs an energy function on the latent variable space that yields an energ…

Low-level Pose Control of Tilting Multirotor for Wall Perching Tasks Using Reinforcement Learning

2021-08-11 · Hyungyu Lee, Myeongwoo Jeong, Chanyoung Kim, Hyungtae Lim 외

Recently, needs for unmanned aerial vehicles (UAVs) that are attachable to the wall have been highlighted. As one of the ways to address the need, researches on various tilting multirotors that can increase maneuverabili…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)

Conformal Bayes for Two-Sided Censored Gaussian Regression under Label Shift

2026-07-02 · Seungjin Choi arxiv

Prediction under label shift becomes nonstandard when responses are censored. In a two-sided censored Gaussian model, latent values below $L$ and above $U$ are recorded at the boundary values, so the observed predictive …