Many standard TTT methods train on carefully selected data from the pre-training dataset (i.e., do not add any new privileged information; Hardt & Sun, 2024; Hübotter et al., 2025), and several works studied how to optimally select data for imitation, e.g., the early seminal work of MacKay (1992) and recent extensions (Hübotter et al., 2024; Bagatella et al., 2025b). TTT has also been extended from supervised learning to reinforcement learning (Zuo et al., 2025; Bagatella et al., 2025a; Diaz-Bone et al., 2025). So far it has not been well understood why and when TTT is effective. While many different methods have been proposed for TTT, we focus here on analyzing "semi-parametric" TTT (e.g., Hardt & Sun, 2024; Hübotter et al., 2025), where a pre-trained model is fine-tuned with a supervised loss on a small neighborhood of the test point in the training data. This is different from some other methods for test-time "adaptation", which are commonly applied with distribution shifts (e.g., Wang et al., 2021; Zhang et al., 2022; Durasov et al., 2025). Basu et al. (2023) consider a similar setting to ours, but analyze it through the lens of non-parametric estimation, relying on the smoothness of the target function in the feature space Ψ.

Accurately predicting protein fitness with minimal experimental data is a persistent challenge in protein engineering. We introduce PRIMO (PRotein In-context Mutation Oracle), a transformer-based framework that leverages in-context learning and test-time training to adapt rapidly to new proteins and assays without large task-specific datasets. By encoding sequence information, auxiliary zero-shot predictions, and sparse experimental labels from many assays as a unified token set in a pre-training masked-language modeling paradigm, PRIMO learns to prioritize promising variants through a preference-based loss function. Across diverse protein families and properties-including both substitution and indel mutations-PRIMO outperforms zero-shot and fully supervised baselines. This work underscores the power of combining large-scale pre-training with efficient test-time adaptation to tackle challenging protein design tasks where data collection is expensive and label availability is limited.

The Abstraction and Reasoning Corpus (ARC) is designed to promote research on abstract reasoning, a fundamental aspect of human intelligence. Common approaches to ARC treat it as a language-oriented problem, addressed by large language models (LLMs) or recurrent reasoning models. However, although the puzzle-like tasks in ARC are inherently visual, existing research has rarely approached the problem from a vision-centric perspective. In this work, we formulate ARC within a vision paradigm, framing it as an image-to-image translation problem. T o incorporate visual priors, we represent the inputs on a "canvas" that can be processed like natural images. It is then natural for us to apply standard vision architectures, such as a vanilla Vision Transformer (ViT), to perform image-to-image mapping. Our model is trained from scratch solely on ARC data and generalizes to unseen tasks through test-time training. Our framework, termed Vision ARC (V ARC), achieves 60.4% accuracy on the ARC-1 benchmark, substantially outperforming existing methods that are also trained from scratch. Our results are competitive with those of leading LLMs and close the gap to average human performance.

The players learn to converge on the "natural" layer decomposition in

Test-time training (TTT) through self-supervised learning (SSL) is an emerging paradigm to tackle distributional shifts. Despite encouraging results, it remains unclear when this approach thrives or fails.

An image is worth 16x16 words: Transformers for image recognition at scale.

Recent advances such as self-consistency and test-time reinforcement learning (TTRL) improve the reliability of large language models (LLMs) without additional supervision, yet their underlying mechanisms and statistical guarantees remain poorly understood. We present a unified framework for certifiable inference in LLMs, showing that majority voting provides a statistical certificate of self-consistency: under mild assumptions, the aggregated answer coincides with the mode of the model's terminal distribution with high probability. We derive finite-sample and anytime-valid concentration bounds that quantify this confidence, and introduce the Martingale Majority Certificate (MMC), a sequential stopping rule that adaptively determines when sufficient samples have been drawn. We further prove that label-free post-training methods such as TTRL implicitly sharpen the answer distribution by exponentially tilting it toward its mode, thereby reducing the number of samples required for certification. Building on this insight, we propose new post-training objectives that explicitly optimise this trade-off between sharpness and bias. Together, these results explain and connect two central test-time scaling strategies, self-consistency and TTRL, within a single statistical framework for label-free, certifiable reliability in reasoning LLMs.

This paper introduces a novel application of Test-Time Training (TTT) for Speech Enhancement, addressing the challenges posed by unpredictable noise conditions and domain shifts. This method combines a main speech enhancement task with a self-supervised auxiliary task in a Y-shaped architecture. The model dynamically adapts to new domains during inference time by optimizing the proposed self-supervised tasks like noise-augmented signal reconstruction or masked spectrogram prediction, bypassing the need for labeled data. We further introduce various TTT strategies offering a trade-off between adaptation and efficiency. Evaluations across synthetic and real-world datasets show consistent improvements across speech quality metrics, outperforming the baseline model. This work highlights the effectiveness of TTT in speech enhancement, providing insights for future research in adaptive and robust speech processing.