What, How, Where, and How Well? A Survey on Test-Time Scaling in Large Language Models

As enthusiasm for scaling computation (data and parameters) in the pretraining era gradually diminished, test-time scaling (TTS), also referred to as ``test-time computing'' has emerged as a prominent research focus. Recent studies demonstrate that TTS can further elicit the problem-solving capabilities of large language models (LLMs), enabling significant breakthroughs not only in specialized reasoning tasks, such as mathematics and coding, but also in general tasks like open-ended Q&A. However, despite the explosion of recent efforts in this area, there remains an urgent need for a comprehensive survey offering a systemic understanding. To fill this gap, we propose a unified, multidimensional framework structured along four core dimensions of TTS research: what to scale, how to scale, where to scale, and how well to scale. Building upon this taxonomy, we conduct an extensive review of methods, application scenarios, and assessment aspects, and present an organized decomposition that highlights the unique functional roles of individual techniques within the broader TTS landscape. From this analysis, we distill the major developmental trajectories of TTS to date and offer hands-on guidelines for practical deployment. Furthermore, we identify several open challenges and offer insights into promising future directions, including further scaling, clarifying the functional essence of techniques, generalizing to more tasks, and more attributions.

Semantic Latent Motion for Portrait Video Generation

Recent advancements in portrait video generation have been noteworthy. However, existing methods rely heavily on human priors and pre-trained generation models, which may introduce unrealistic motion and lead to inefficient inference. To address these challenges, we propose Semantic Latent Motion (SeMo), a compact and expressive motion representation. Leveraging this representation, our approach achieve both high-quality visual results and efficient inference. SeMo follows an effective three-step framework: Abstraction, Reasoning, and Generation. First, in the Abstraction step, we use a carefully designed Mask Motion Encoder to compress the subject's motion state into a compact and abstract latent motion (1D token). Second, in the Reasoning step, long-term modeling and efficient reasoning are performed in this latent space to generate motion sequences. Finally, in the Generation step, the motion dynamics serve as conditional information to guide the generation model in synthesizing realistic transitions from reference frames to target frames. Thanks to the compact and descriptive nature of Semantic Latent Motion, our method enables real-time video generation with highly realistic motion. User studies demonstrate that our approach surpasses state-of-the-art models with an 81% win rate in realism. Extensive experiments further highlight its strong compression capability, reconstruction quality, and generative potential. Moreover, its fully self-supervised nature suggests promising applications in broader video generation tasks.

Crowd Comparative Reasoning: Unlocking Comprehensive Evaluations for LLM-as-a-Judge

LLM-as-a-Judge, which generates chain-of-thought (CoT) judgments, has become a widely adopted auto-evaluation method. However, its reliability is compromised by the CoT reasoning's inability to capture comprehensive and deeper details, often leading to incomplete outcomes. Existing methods mainly rely on majority voting or criteria expansion, which is insufficient to address the limitation in CoT. We propose Crowd-based Comparative Evaluation, which introduces additional crowd responses to compare with the candidate responses, thereby exposing deeper and more comprehensive details within the candidate responses. This process effectively guides LLM-as-a-Judge to provide a more detailed CoT judgment. Extensive experiments demonstrate that our approach enhances evaluation reliability, achieving an average accuracy gain of 6.7% across five benchmarks. Moreover, our method produces higher-quality CoTs that facilitate judge distillation and exhibit superior performance in rejection sampling for supervised fine-tuning (SFT), referred to as crowd rejection sampling, thereby enabling more efficient SFT. Our analysis confirms that CoTs generated by ours are more comprehensive and of higher quality, and evaluation accuracy improves as inference scales.

HiLo: Learning Whole-Body Human-like Locomotion with Motion Tracking Controller

Deep Reinforcement Learning (RL) has emerged as a promising method to develop humanoid robot locomotion controllers. Despite the robust and stable locomotion demonstrated by previous RL controllers, their behavior often lacks the natural and agile motion patterns necessary for human-centric scenarios. In this work, we propose HiLo (human-like locomotion with motion tracking), an effective framework designed to learn RL policies that perform human-like locomotion. The primary challenges of human-like locomotion are complex reward engineering and domain randomization. HiLo overcomes these issues by developing an RL-based motion tracking controller and simple domain randomization through random force injection and action delay. Within the framework of HiLo, the whole-body control problem can be decomposed into two components: One part is solved using an open-loop control method, while the residual part is addressed with RL policies. A distributional value function is also implemented to stabilize the training process by improving the estimation of cumulative rewards under perturbed dynamics. Our experiments demonstrate that the motion tracking controller trained using HiLo can perform natural and agile human-like locomotion while exhibiting resilience to external disturbances in real-world systems. Furthermore, we show that the motion patterns of humanoid robots can be adapted through the residual mechanism without fine-tuning, allowing quick adjustments to task requirements.

UniAvatar: Taming Lifelike Audio-Driven Talking Head Generation with Comprehensive Motion and Lighting Control

Recently, animating portrait images using audio input is a popular task. Creating lifelike talking head videos requires flexible and natural movements, including facial and head dynamics, camera motion, realistic light and shadow effects. Existing methods struggle to offer comprehensive, multifaceted control over these aspects. In this work, we introduce UniAvatar, a designed method that provides extensive control over a wide range of motion and illumination conditions. Specifically, we use the FLAME model to render all motion information onto a single image, maintaining the integrity of 3D motion details while enabling fine-grained, pixel-level control. Beyond motion, this approach also allows for comprehensive global illumination control. We design independent modules to manage both 3D motion and illumination, permitting separate and combined control. Extensive experiments demonstrate that our method outperforms others in both broad-range motion control and lighting control. Additionally, to enhance the diversity of motion and environmental contexts in current datasets, we collect and plan to publicly release two datasets, DH-FaceDrasMvVid-100 and DH-FaceReliVid-200, which capture significant head movements during speech and various lighting scenarios.

NILE: Internal Consistency Alignment in Large Language Models

As a crucial step to enhance LLMs alignment with human intentions, Instruction Fine-Tuning (IFT) has a high demand on dataset quality. However, existing IFT datasets often contain knowledge that is inconsistent with LLMs' internal knowledge learned from the pre-training phase, which can greatly affect the efficacy of IFT. To address this issue, we introduce NILE (iNternal consIstency aLignmEnt) framework, aimed at optimizing IFT datasets to unlock LLMs' capability further. NILE operates by eliciting target pre-trained LLM's internal knowledge corresponding to instruction data. The internal knowledge is leveraged to revise the answer in IFT datasets. Additionally, we propose a novel Internal Consistency Filtering (ICF) method to filter training samples, ensuring its high consistency with LLM's internal knowledge. Our experiments demonstrate that NILE-aligned IFT datasets sharply boost LLM performance across multiple LLM ability evaluation datasets, achieving up to 66.6% gain on Arena-Hard and 68.5% on Alpaca-Eval V2. Further analysis confirms that each component of the NILE}framework contributes to these substantial performance improvements, and provides compelling evidence that dataset consistency with pre-trained internal knowledge is pivotal for maximizing LLM potential.

RevisEval: Improving LLM-as-a-Judge via Response-Adapted References

With significant efforts in recent studies, LLM-as-a-Judge has become a cost-effective alternative to human evaluation for assessing the text generation quality in a wide range of tasks. However, there still remains a reliability gap between LLM-as-a-Judge and human evaluation. One important reason is the lack of guided oracles in the evaluation process. Motivated by the role of reference pervasively used in classic text evaluation, we introduce RevisEval, a novel text generation evaluation paradigm via the response-adapted references. RevisEval is driven by the key observation that an ideal reference should maintain the necessary relevance to the response to be evaluated. Specifically, RevisEval leverages the text revision capabilities of large language models (LLMs) to adaptively revise the response, then treat the revised text as the reference (response-adapted reference) for the subsequent evaluation. Extensive experiments demonstrate that RevisEval outperforms traditional reference-free and reference-based evaluation paradigms that use LLM-as-a-Judge across NLG tasks and open-ended instruction-following tasks. More importantly, our response-adapted references can further boost the classical text metrics, e.g., BLEU and BERTScore, compared to traditional references and even rival the LLM-as-a-Judge. A detailed analysis is also conducted to confirm RevisEval's effectiveness in bias reduction, the impact of inference cost, and reference relevance.

Collaborative Performance Prediction for Large Language Models

Comprehensively understanding and accurately predicting the performance of large language models across diverse downstream tasks has emerged as a pivotal challenge in NLP research. The pioneering scaling law on downstream works demonstrated intrinsic similarities within model families and utilized such similarities for performance prediction. However, they tend to overlook the similarities between model families and only consider design factors listed in the original scaling law. To overcome these limitations, we introduce a novel framework, Collaborative Performance Prediction (CPP), which significantly enhances prediction accuracy by leveraging the historical performance of various models on downstream tasks and other design factors for both model and task. We also collect a collaborative data sourced from online platforms containing both historical performance and additional design factors. With the support of the collaborative data, CPP not only surpasses traditional scaling laws in predicting the performance of scaled LLMs but also facilitates a detailed analysis of factor importance, an area previously overlooked.

NOAHQA: Numerical Reasoning with Interpretable Graph Question Answering Dataset

While diverse question answering (QA) datasets have been proposed and contributed significantly to the development of deep learning models for QA tasks, the existing datasets fall short in two aspects. First, we lack QA datasets covering complex questions that involve answers as well as the reasoning processes to get the answers. As a result, the state-of-the-art QA research on numerical reasoning still focuses on simple calculations and does not provide the mathematical expressions or evidences justifying the answers. Second, the QA community has contributed much effort to improving the interpretability of QA models. However, these models fail to explicitly show the reasoning process, such as the evidence order for reasoning and the interactions between different pieces of evidence. To address the above shortcomings, we introduce NOAHQA, a conversational and bilingual QA dataset with questions requiring numerical reasoning with compound mathematical expressions. With NOAHQA, we develop an interpretable reasoning graph as well as the appropriate evaluation metric to measure the answer quality. We evaluate the state-of-the-art QA models trained using existing QA datasets on NOAHQA and show that the best among them can only achieve 55.5 exact match scores, while the human performance is 89.7. We also present a new QA model for generating a reasoning graph where the reasoning graph metric still has a large gap compared with that of humans, e.g., 28 scores.

MWPToolkit: An Open-Source Framework for Deep Learning-Based Math Word Problem Solvers

Developing automatic Math Word Problem (MWP) solvers has been an interest of NLP researchers since the 1960s. Over the last few years, there are a growing number of datasets and deep learning-based methods proposed for effectively solving MWPs. However, most existing methods are benchmarked soly on one or two datasets, varying in different configurations, which leads to a lack of unified, standardized, fair, and comprehensive comparison between methods. This paper presents MWPToolkit, the first open-source framework for solving MWPs. In MWPToolkit, we decompose the procedure of existing MWP solvers into multiple core components and decouple their models into highly reusable modules. We also provide a hyper-parameter search function to boost the performance. In total, we implement and compare 17 MWP solvers on 4 widely-used single equation generation benchmarks and 2 multiple equations generation benchmarks. These features enable our MWPToolkit to be suitable for researchers to reproduce advanced baseline models and develop new MWP solvers quickly. Code and documents are available at https://github.com/LYH-YF/MWPToolkit.