Fine-Tuning

Fine-tuning has become the dominant paradigm for adapting Vision-Language Models (VLMs), yet most approaches rely on explicit weight updates that introduce a fundamental trade-off. Full Fine-Tuning (FFT) may perturb pretrained representations due to cross-modal gradient interference, whereas Parameter-Efficient Fine-Tuning (PEFT) methods rely on additive modules, such as low-rank adapters, which may limit adaptation capacity. In this paper, we rethink VLM adaptation from a structural selection framework that adapts VLMs without modifying backbone weights, and we propose Mask Fine-Tuning (MFT). MFT learns masks that selectively route information through existing pretrained connections, dynamically uncovering subnetworks that better align pretrained representations with downstream objectives. Extensive experiments show that MFT provides an effective structural alternative to both FFT and PEFT, consistently achieving superior performance across multiple vision-language benchmarks without adding knowledge or altering the deployment architecture. Moreover, our analysis with MFT provides new insights into how pretrained VLMs reorganize their internal representational pathways during adaptation.

Unequal representation of demographic groups in training data poses challenges to model generalisation across populations. Standard practice assumes that balancing subgroup representation always optimises performance. However, recent empirical results contradict this assumption: in some cases, imbalanced data distributions actually improve subgroup performance, and in others, subgroup performance is unaffected by the absence of an entire subgroup during training. We propose the latent separation hypothesis, which states that a fine-tuned model's dependence on subgroup representation is determined by the degree of separation between subgroups in the latent space of the pre-trained model. We formalise this hypothesis, provide theoretical analysis, validate it empirically across four vision and language datasets, and present preliminary evidence of a causal role through regularisation experiments. Finally, we show how quantitative analysis of latent subgroup separation can inform practical data collection and balancing decisions when fine-tuning foundation models.

Language model watermarking schemes fall into two broad categories: inference-time methods, which modify the decoding process at generation time, and model-embedded methods, which encode a secret signal directly into the model weights. Inference-time watermarking methods incur additional inference overhead and are not applicable in the increasingly prevalent open-weight model settings. In contrast, model-embedded watermarks introduce no additional inference latency-generation uses the standard sampling pipeline-and are particularly well-suited to open-weight settings. However, existing model-embedded approaches such as GaussMark face a fundamental quality-detectability trade-off: achieving strong detection power typically requires weight perturbations that noticeably degrade generation quality. Our key insight is that regularizing toward a high-quality reference distribution decouples quality preservation from proximity to the base model, enabling a better quality-detectability trade-off. Building on this insight, we introduce MarkTune, a theoretically grounded on-policy fine-tuning framework that uses the GaussMark detection statistic as a reward while preserving generation quality through cross-entropy on high-quality human text or KL divergence to a stronger oracle model. Empirically, MarkTune substantially improves the quality-detectability frontier of GaussMark, approaching the detectability of strong inference-time schemes while preserving generation quality and downstream task performance. Marktune also remains robust under the evaluated text-editing and open-weight attacks and generalizes across datasets, retaining substantial detection power on corpora unseen during training. Code is available at https://github.com/zhyzmath/MarkTune-LLM-Watermarking.

Diffusion models have achieved remarkable success in generative modeling, yet how to effectively adapt large pretrained models to new tasks remains challenging. We revisit the reconstruction behavior of diffusion models during denoising to unveil the underlying frequency energy mechanism governing this process. Building upon this observation, we propose FeRA, a frequency driven fine tuning framework that aligns parameter updates with the intrinsic frequency energy progression of diffusion. FeRA establishes a comprehensive frequency energy framework for effective diffusion adaptation fine tuning, comprising three synergistic components: (i) a compact frequency energy indicator that characterizes the latent bandwise energy distribution, (ii) a soft frequency router that adaptively fuses multiple frequency specific adapter experts, and (iii) a frequency energy consistency regularization that stabilizes diffusion optimization and ensures coherent adaptation across bands. Routing operates in both training and inference, with inference time routing dynamically determined by the latent frequency energy. It integrates seamlessly with adapter based tuning schemes and generalizes well across diffusion backbones and resolutions. By aligning adaptation with the frequency energy mechanism, FeRA provides a simple, stable, and compatible paradigm for effective and robust diffusion model adaptation.

Fine-tuning large language models (LLMs) for downstream tasks is an essential stage of modern AI deployment. Reinforcement learning (RL) has emerged as the dominant fine-tuning paradigm, underpinning many state-of-the-art LLMs. In contrast, evolution strategies (ES) has largely been overlooked due to the widespread belief that it does not scale to modern model sizes. This paper overturns this assumption by demonstrating the first successful application of ES to full-parameter fine-tuning of LLMs at the billion-parameter scale, without dimensionality reduction. ES can indeed search over extremely high-dimensional parameter spaces and outperform established RL implementations across multiple axes, including improved tolerance to long-horizon and delayed rewards, robustness across diverse base LLMs, reduced susceptibility to reward hacking, and improved training stability. These findings suggest that ES is not merely a viable alternative to RL, but a fundamentally different and powerful backpropagation-free post-training paradigm that opens a new direction for LLM fine-tuning beyond current RL-based approaches.

Transformers often transfer across related tasks with little fine-tuning; we study this question in noisy nonparametric regression for continuous maps $f,g:[0,1]^d\to\mathbb{R}^D$, representing the source and shifted tasks, using a single-query, single-head transformer $\mathcal{T}(\cdot)=V\operatorname{Softmax}(-α(\cdot,K))$ with $\ell^{\infty}$-based alignment scores $α$. Rather than assume that gradient-based training reaches an empirical risk minimizer, we give a model-specific two-stage procedure certifying source-task recovery and out-of-distribution (OOD) adaptation. With high probability, the resulting attention head recovers $f$ to arbitrary accuracy $\varepsilon>0$ in every fixed $L^p([0,1]^d)$ norm, $1\le p<\infty$, from noisy observations. For Lipschitz targets, its rate is minimax-optimal up to logarithmic factors without assuming oracle access to an asymptotic ERM. If $D=1$ and $f$ is Pfaffian than the values matrix factors into $AC$ with $A$ scaling at-most linearly $\mathcal{O}(\varepsilon^{-1})$ in the reciprocal approximation error $\varepsilon>0$. Importantly, our training procedure encodes all our identified (potentially) shared geometric structure from $f$ only into the values matrix factor $C$ and the keys matrix $K$ allowing for rapid access, and all $f$-specific structure only into $A$. Our main result concerns the OOD phase; where we identify geometric structures, which if shared by the source $f$ and OOD $g$ task; then only only the values matrix factor $A$ needs to be fine-tuned to guarantee an $L^p$-recovery of $g$. Critically, fine-tuning only $A$ is a convex problem which involving an at-most linear number of parameters $\mathcal{O}(\varepsilon^{-1})$; dramatically cutting down on the worst-case scaling of $\mathcal{O}(\varepsilon^{-\mathbf{d}})$ unavoidable in the source-training phase.

Semi-supervised domain adaptation (SSDA) seeks to achieve accurate predictions in a target domain with limited labeled target data by exploiting abundant source and unlabeled target data. We study this problem under structural causal models (SCMs), which provide a statistical framework to describe distribution shifts between source and target domains as interventions in the data-generating process rather than ad hoc changes in model parameters. The central phenomenon is that, under low-dimensional interventions, source and unlabeled target data can help identify the high-dimensional shared structure, leaving only a low-dimensional target-specific correction to be learned from limited labeled target data. We formalize this principle for three canonical intervention models and propose the corresponding SSDA methods FT-DIP, FT-OLS-Src and FT-CIP. Under each intervention model, we demonstrate how extending an unsupervised domain adaptation (UDA) method to SSDA can achieve minimax-optimal target performance with limited target labels, with the labeled-target sample complexity scaling with the intervention dimension rather than the ambient dimension. When the distribution shift is underspecified, we propose the Multi-Adaptive-Start Fine-Tuning (MASFT) algorithm, which fine-tunes from multiple adaptive starts and selects among them using a small target validation set, incurring only logarithmic overhead in the number of starts. We validate the effectiveness of our proposed methods through simulated and real data experiments.

Dynamic Scene Graph Generation (DSGG) aims to capture objects and their evolving relations in videos. Despite recent progress, the practicality and quality of generated scene graphs remain limited compared to the rapid advances in Multimodal Large Language Models (MLLMs). In this work, we revisit DSGG from two fundamental perspectives: task setup and model design. From the task setup perspective, we identify two key limitations of the current recall-oriented evaluation protocol: (i) a severe precision-recall trade-off, and (ii) uninformative and redundant relation generation. To better assess practical usefulness, we introduce five additional metrics that provide a more comprehensive evaluation of the quality of generated dynamic scene graphs. From the model design perspective, we explore directly using MLLMs for scene graph generation and establish a strong MLLM-based DSGG baseline through three design changes. First, we replace the conventional bottom-up pipeline with a top-down reason-then-locate strategy. Second, we reformulate frame-wise dynamic graphs as Temporal Relation Set (TRS) prediction, improving both efficiency and performance. Third, we introduce Importance-Aware Finetuning (IAF) to encourage more relevant and diverse relation generation. Extensive experiments on Action Genome, VidVRD, and PVSG show that our approach consistently achieves state-of-the-art performance.

Transfer Learning (TL) is an emerging field in modeling building thermal dynamics. This method reduces the data required for a data-driven model of a target building by leveraging knowledge from a source building. Consequently, it enables the creation of data-efficient models that can be used for advanced control and fault detection & diagnosis. A major limitation of the TL approach is its inconsistent performance across different sources. Although accurate source-building selection for a target is crucial, it remains a persistent challenge. We present GenTL, a general transfer learning model for single-family houses in Central Europe. GenTL can be efficiently fine-tuned to a large variety of target buildings. It is pretrained on a Long Short-Term Memory (LSTM) network with data from 450 different buildings. The general transfer learning model eliminates the need for source-building selection by serving as a universal source for fine-tuning. Comparative analysis with conventional single-source to single-target TL demonstrates the efficacy and reliability of the general pretraining approach. Testing GenTL on 144 target buildings for fine-tuning reveals an average prediction error (RMSE) reduction of 42.1 % compared to fine-tuning single-source models.

In out-of-distribution (OOD) detection, fine-tuning with auxiliary outlier data often improves detection performance at the cost of classification accuracy. This trade-off stems from the loss of the original in-distribution (ID) distribution during fine-tuning. To establish a more practical and effective paradigm, we optimize three critical factors: model reminder, data sampling, and representation learning. We propose: (1) Self-Knowledge Distillation (SKD) to mitigate accuracy reduction; (2) Semi-hard Outlier Sampling (SOS) to improve detection efficiency with minimal data; and (3) Outlier-aware Supervised Contrastive Learning (OSCL) to promote ID-OOD separability. Optimizing these factors produces cumulative gains, boosting both OOD detection performance and classification accuracy. Our framework outperforms existing methods across diverse benchmarks, particularly in long-tailed scenarios, providing a robust baseline for real-world OOD detection.