Revisiting LLM Evaluation through Mechanism Interpretability: a New Metric and Model Utility Law

Large Language Models (LLMs) have become indispensable across academia, industry, and daily applications, yet current evaluation methods struggle to keep pace with their rapid development. In this paper, we analyze the core limitations of traditional evaluation pipelines and propose a novel metric, the Model Utilization Index (MUI), which introduces mechanism interpretability techniques to complement traditional performance metrics. MUI quantifies the extent to which a model leverages its capabilities to complete tasks. The core idea is that to assess an LLM's overall ability, we must evaluate not only its task performance but also the effort expended to achieve the outcome. Our extensive experiments reveal an inverse relationship between MUI and performance, from which we deduce a common trend observed in popular LLMs, which we term the Utility Law. Based on this, we derive four corollaries that address key challenges, including training judgement, the issue of data contamination, fairness in model comparison, and data diversity. We hope that our survey, novel metric, and utility law will foster mutual advancement in both evaluation and mechanism interpretability. Our code can be found at https://github.com/ALEX-nlp/MUI-Eva.

FamilyTool: A Multi-hop Personalized Tool Use Benchmark

The integration of tool learning with Large Language Models (LLMs) has expanded their capabilities in handling complex tasks by leveraging external tools. However, existing benchmarks for tool learning inadequately address critical real-world personalized scenarios, particularly those requiring multi-hop reasoning and inductive knowledge adaptation in dynamic environments. To bridge this gap, we introduce FamilyTool, a novel benchmark grounded in a family-based knowledge graph (KG) that simulates personalized, multi-hop tool use scenarios. FamilyTool challenges LLMs with queries spanning 1 to 3 relational hops (e.g., inferring familial connections and preferences) and incorporates an inductive KG setting where models must adapt to unseen user preferences and relationships without re-training, a common limitation in prior approaches that compromises generalization. We further propose KGETool: a simple KG-augmented evaluation pipeline to systematically assess LLMs' tool use ability in these settings. Experiments reveal significant performance gaps in state-of-the-art LLMs, with accuracy dropping sharply as hop complexity increases and inductive scenarios exposing severe generalization deficits. These findings underscore the limitations of current LLMs in handling personalized, evolving real-world contexts and highlight the urgent need for advancements in tool-learning frameworks. FamilyTool serves as a critical resource for evaluating and advancing LLM agents' reasoning, adaptability, and scalability in complex, dynamic environments. Code and dataset are available at Github.

World Modeling Makes a Better Planner: Dual Preference Optimization for Embodied Task Planning

Recent advances in large vision-language models (LVLMs) have shown promise for embodied task planning, yet they struggle with fundamental challenges like dependency constraints and efficiency. Existing approaches either solely optimize action selection or leverage world models during inference, overlooking the benefits of learning to model the world as a way to enhance planning capabilities. We propose Dual Preference Optimization (D$^2$PO), a new learning framework that jointly optimizes state prediction and action selection through preference learning, enabling LVLMs to understand environment dynamics for better planning. To automatically collect trajectories and stepwise preference data without human annotation, we introduce a tree search mechanism for extensive exploration via trial-and-error. Extensive experiments on VoTa-Bench demonstrate that our D$^2$PO-based method significantly outperforms existing methods and GPT-4o when applied to Qwen2-VL (7B), LLaVA-1.6 (7B), and LLaMA-3.2 (11B), achieving superior task success rates with more efficient execution paths.

YuE: Scaling Open Foundation Models for Long-Form Music Generation

We tackle the task of long-form music generation--particularly the challenging \textbf{lyrics-to-song} problem--by introducing YuE, a family of open foundation models based on the LLaMA2 architecture. Specifically, YuE scales to trillions of tokens and generates up to five minutes of music while maintaining lyrical alignment, coherent musical structure, and engaging vocal melodies with appropriate accompaniment. It achieves this through (1) track-decoupled next-token prediction to overcome dense mixture signals, (2) structural progressive conditioning for long-context lyrical alignment, and (3) a multitask, multiphase pre-training recipe to converge and generalize. In addition, we redesign the in-context learning technique for music generation, enabling versatile style transfer (e.g., converting Japanese city pop into an English rap while preserving the original accompaniment) and bidirectional generation. Through extensive evaluation, we demonstrate that YuE matches or even surpasses some of the proprietary systems in musicality and vocal agility. In addition, fine-tuning YuE enables additional controls and enhanced support for tail languages. Furthermore, beyond generation, we show that YuE's learned representations can perform well on music understanding tasks, where the results of YuE match or exceed state-of-the-art methods on the MARBLE benchmark. Keywords: lyrics2song, song generation, long-form, foundation model, music generation

How to Mitigate Overfitting in Weak-to-strong Generalization?

Aligning powerful AI models on tasks that surpass human evaluation capabilities is the central problem of \textbf{superalignment}. To address this problem, weak-to-strong generalization aims to elicit the capabilities of strong models through weak supervisors and ensure that the behavior of strong models aligns with the intentions of weak supervisors without unsafe behaviors such as deception. Although weak-to-strong generalization exhibiting certain generalization capabilities, strong models exhibit significant overfitting in weak-to-strong generalization: Due to the strong fit ability of strong models, erroneous labels from weak supervisors may lead to overfitting in strong models. In addition, simply filtering out incorrect labels may lead to a degeneration in question quality, resulting in a weak generalization ability of strong models on hard questions. To mitigate overfitting in weak-to-strong generalization, we propose a two-stage framework that simultaneously improves the quality of supervision signals and the quality of input questions. Experimental results in three series of large language models and two mathematical benchmarks demonstrate that our framework significantly improves PGR compared to naive weak-to-strong generalization, even achieving up to 100\% PGR on some models.

Prior-Fitted Networks Scale to Larger Datasets When Treated as Weak Learners

Prior-Fitted Networks (PFNs) have recently been proposed to efficiently perform tabular classification tasks. Although they achieve good performance on small datasets, they encounter limitations with larger datasets. These limitations include significant memory consumption and increased computational complexity, primarily due to the impracticality of incorporating all training samples as inputs within these networks. To address these challenges, we investigate the fitting assumption for PFNs and input samples. Building on this understanding, we propose \textit{BoostPFN} designed to enhance the performance of these networks, especially for large-scale datasets. We also theoretically validate the convergence of BoostPFN and our empirical results demonstrate that the BoostPFN method can outperform standard PFNs with the same size of training samples in large datasets and achieve a significant acceleration in training times compared to other established baselines in the field, including widely-used Gradient Boosting Decision Trees (GBDTs), deep learning methods and AutoML systems. High performance is maintained for up to 50x of the pre-training size of PFNs, substantially extending the limit of training samples. Through this work, we address the challenges of efficiently handling large datasets via PFN-based models, paving the way for faster and more effective tabular data classification training and prediction process. Code is available at Github.

CritiQ: Mining Data Quality Criteria from Human Preferences

Language model heavily depends on high-quality data for optimal performance. Existing approaches rely on manually designed heuristics, the perplexity of existing models, training classifiers, or careful prompt engineering, which require significant expert experience and human annotation effort while introduce biases. We introduce CritiQ, a novel data selection method that automatically mines criteria from human preferences for data quality with only $\sim$30 human-annotated pairs and performs efficient data selection. The main component, CritiQ Flow, employs a manager agent to evolve quality criteria and worker agents to make pairwise judgments. We build a knowledge base that extracts quality criteria from previous work to boost CritiQ Flow. Compared to perplexity- and classifier- based methods, verbal criteria are more interpretable and possess reusable value. After deriving the criteria, we train the CritiQ Scorer to give quality scores and perform efficient data selection. We demonstrate the effectiveness of our method in the code, math, and logic domains, achieving high accuracy on human-annotated test sets. To validate the quality of the selected data, we continually train Llama 3.1 models and observe improved performance on downstream tasks compared to uniform sampling. Ablation studies validate the benefits of the knowledge base and the reflection process. We analyze how criteria evolve and the effectiveness of majority voting.

Thus Spake Long-Context Large Language Model

Long context is an important topic in Natural Language Processing (NLP), running through the development of NLP architectures, and offers immense opportunities for Large Language Models (LLMs) giving LLMs the lifelong learning potential akin to humans. Unfortunately, the pursuit of a long context is accompanied by numerous obstacles. Nevertheless, long context remains a core competitive advantage for LLMs. In the past two years, the context length of LLMs has achieved a breakthrough extension to millions of tokens. Moreover, the research on long-context LLMs has expanded from length extrapolation to a comprehensive focus on architecture, infrastructure, training, and evaluation technologies. Inspired by the symphonic poem, Thus Spake Zarathustra, we draw an analogy between the journey of extending the context of LLM and the attempts of humans to transcend its mortality. In this survey, We will illustrate how LLM struggles between the tremendous need for a longer context and its equal need to accept the fact that it is ultimately finite. To achieve this, we give a global picture of the lifecycle of long-context LLMs from four perspectives: architecture, infrastructure, training, and evaluation, showcasing the full spectrum of long-context technologies. At the end of this survey, we will present 10 unanswered questions currently faced by long-context LLMs. We hope this survey can serve as a systematic introduction to the research on long-context LLMs.

AutoLogi: Automated Generation of Logic Puzzles for Evaluating Reasoning Abilities of Large Language Models

While logical reasoning evaluation of Large Language Models (LLMs) has attracted significant attention, existing benchmarks predominantly rely on multiple-choice formats that are vulnerable to random guessing, leading to overestimated performance and substantial performance fluctuations. To obtain more accurate assessments of models' reasoning capabilities, we propose an automated method for synthesizing open-ended logic puzzles, and use it to develop a bilingual benchmark, AutoLogi. Our approach features program-based verification and controllable difficulty levels, enabling more reliable evaluation that better distinguishes models' reasoning abilities. Extensive evaluation of eight modern LLMs shows that AutoLogi can better reflect true model capabilities, with performance scores spanning from 35% to 73% compared to the narrower range of 21% to 37% on the source multiple-choice dataset. Beyond benchmark creation, this synthesis method can generate high-quality training data by incorporating program verifiers into the rejection sampling process, enabling systematic enhancement of LLMs' reasoning capabilities across diverse datasets.

Human2Robot: Learning Robot Actions from Paired Human-Robot Videos

Distilling knowledge from human demonstrations is a promising way for robots to learn and act. Existing work often overlooks the differences between humans and robots, producing unsatisfactory results. In this paper, we study how perfectly aligned human-robot pairs benefit robot learning. Capitalizing on VR-based teleportation, we introduce H\&R, a third-person dataset with 2,600 episodes, each of which captures the fine-grained correspondence between human hands and robot gripper. Inspired by the recent success of diffusion models, we introduce Human2Robot, an end-to-end diffusion framework that formulates learning from human demonstrates as a generative task. Human2Robot fully explores temporal dynamics in human videos to generate robot videos and predict actions at the same time. Through comprehensive evaluations of 8 seen, changed and unseen tasks in real-world settings, we demonstrate that Human2Robot can not only generate high-quality robot videos but also excel in seen tasks and generalize to unseen objects, backgrounds and even new tasks effortlessly.