Jie Ruan

Mingqi Gao, Jie Ruan, Renliang Sun et al.

Evaluating text summarization is a challenging problem, and existing evaluation metrics are far from satisfactory. In this study, we explored ChatGPT's ability to perform human-like summarization evaluation using four human evaluation methods on five datasets. We found that ChatGPT was able to complete annotations relatively smoothly using Likert scale scoring, pairwise comparison, Pyramid, and binary factuality evaluation. Additionally, it outperformed commonly used automatic evaluation metrics on some datasets. Furthermore, we discussed the impact of different prompts, compared its performance with that of human evaluation, and analyzed the generated explanations and invalid responses.

Jie Ruan, Yue Wu, Xiaojun Wan et al.

Sarcasm generation has been investigated in previous studies by considering it as a text-to-text generation problem, i.e., generating a sarcastic sentence for an input sentence. In this paper, we study a new problem of cross-modal sarcasm generation (CMSG), i.e., generating a sarcastic description for a given image. CMSG is challenging as models need to satisfy the characteristics of sarcasm, as well as the correlation between different modalities. In addition, there should be some inconsistency between the two modalities, which requires imagination. Moreover, high-quality training data is insufficient. To address these problems, we take a step toward generating sarcastic descriptions from images without paired training data and propose an Extraction-Generation-Ranking based Modular method (EGRM) for cross-model sarcasm generation. Specifically, EGRM first extracts diverse information from an image at different levels and uses the obtained image tags, sentimental descriptive caption, and commonsense-based consequence to generate candidate sarcastic texts. Then, a comprehensive ranking algorithm, which considers image-text relation, sarcasticness, and grammaticality, is proposed to select a final text from the candidate texts. Human evaluation at five criteria on a total of 1200 generated image-text pairs from eight systems and auxiliary automatic evaluation show the superiority of our method.

Inderjeet Nair, Jie Ruan, Lu Wang

Alignment faking, where a model behaves aligned with developer policy when monitored but reverts to its own preferences when unobserved, is a concerning yet poorly understood phenomenon, in part because current diagnostic tools remain limited. Prior diagnostics rely on highly toxic and clearly harmful scenarios, causing most models to refuse immediately. As a result, models never deliberate over developer policy, monitoring conditions, or the consequences of non-compliance, making these diagnostics fundamentally unable to detect alignment faking propensity. To support study of this phenomenon, we first introduce VLAF, a diagnostic framework grounded in the hypothesis that alignment faking is most likely when developer policy conflicts with a model's strongly held values. VLAF uses morally unambiguous scenarios to probe this conflict across diverse moral values, bypassing refusal behavior while preserving meaningful deliberative stakes. Using VLAF, we find that alignment faking is substantially more prevalent than previously reported, occurring in models as small as 7B parameters - with olmo2-7b-instruct faking alignment in 37% of cases.Finally, we show that oversight conditions induce activation shifts that lie along a single direction in representation space. This means the behavioral divergence driving alignment faking can be captured by a single contrastive steering vector, which we exploit for lightweight inference-time mitigation. Finally, we exploit this for mitigation that requires no labeled data and minimal computational overhead, achieving relative reductions in alignment faking of 85.8%, 94.0%, and 57.7% on olmo2-7b-instruct, olmo2-13b-instruct, and qwen3-8b respectively.

Mingqi Gao, Xinyu Hu, Jie Ruan et al. · pku

Evaluating natural language generation (NLG) is a vital but challenging problem in natural language processing. Traditional evaluation metrics mainly capturing content (e.g. n-gram) overlap between system outputs and references are far from satisfactory, and large language models (LLMs) such as ChatGPT have demonstrated great potential in NLG evaluation in recent years. Various automatic evaluation methods based on LLMs have been proposed, including metrics derived from LLMs, prompting LLMs, fine-tuning LLMs, and human-LLM collaborative evaluation. In this survey, we first give a taxonomy of LLM-based NLG evaluation methods, and discuss their pros and cons, respectively. Lastly, we discuss several open problems in this area and point out future research directions.

Xunjian Yin, Xu Zhang, Jie Ruan et al. · pku

In recent years, substantial advancements have been made in the development of large language models, achieving remarkable performance across diverse tasks. To evaluate the knowledge ability of language models, previous studies have proposed lots of benchmarks based on question-answering pairs. We argue that it is not reliable and comprehensive to evaluate language models with a fixed question or limited paraphrases as the query, since language models are sensitive to prompt. Therefore, we introduce a novel concept named knowledge boundary to encompass both prompt-agnostic and prompt-sensitive knowledge within language models. Knowledge boundary avoids prompt sensitivity in language model evaluations, rendering them more dependable and robust. To explore the knowledge boundary for a given model, we propose projected gradient descent method with semantic constraints, a new algorithm designed to identify the optimal prompt for each piece of knowledge. Experiments demonstrate a superior performance of our algorithm in computing the knowledge boundary compared to existing methods. Furthermore, we evaluate the ability of multiple language models in several domains with knowledge boundary.

Jie Ruan, Inderjeet Nair, Shuyang Cao et al.

This paper introduces ExpertLongBench, an expert-level benchmark containing 11 tasks from 9 domains that reflect realistic expert workflows and applications. Beyond question answering, the application-driven tasks in ExpertLongBench demand long-form outputs that can exceed 5,000 tokens and strict adherence to domain-specific requirements. Notably, each task in ExpertLongBench includes a rubric, designed or validated by domain experts, to specify task requirements and guide output evaluation. Furthermore, we propose CLEAR, an evaluation framework that supports accurate evaluation of long-form model outputs in our benchmark. To achieve fine-grained, expert-aligned evaluation, CLEAR derives checklists from both model outputs and references by extracting information corresponding to items in the task-specific rubric. Checklist items of model outputs are then compared with corresponding items of reference outputs to assess their correctness, enabling grounded evaluation. We benchmark 13 popular large language models (LLMs) and analyze components in CLEAR, showing that (1) existing LLMs, with the top performer Gemini-2.5-Pro achieving only a 33.4 F1 score, require significant improvement for expert-level tasks; (2) models can generate content corresponding to the required aspects, but far from correct; and (3) accurate checklist extraction and comparison in CLEAR can be achieved by open-weight models for more scalable, reproducible, and low-cost usage.

Jie Ruan, Xiao Pu, Mingqi Gao et al.

Human evaluation is viewed as a reliable evaluation method for NLG which is expensive and time-consuming. To save labor and costs, researchers usually perform human evaluation on a small subset of data sampled from the whole dataset in practice. However, different selection subsets will lead to different rankings of the systems. To give a more correct inter-system ranking and make the gold standard human evaluation more reliable, we propose a Constrained Active Sampling Framework (CASF) for reliable human judgment. CASF operates through a Learner, a Systematic Sampler and a Constrained Controller to select representative samples for getting a more correct inter-system ranking.Experiment results on 137 real NLG evaluation setups with 44 human evaluation metrics across 16 datasets and 5 NLG tasks demonstrate CASF receives 93.18% top-ranked system recognition accuracy and ranks first or ranks second on 90.91% of the human metrics with 0.83 overall inter-system ranking Kendall correlation.Code and data are publicly available online.

Jie Ruan, Wenqing Wang, Xiaojun Wan

Human evaluation serves as the gold standard for assessing the quality of Natural Language Generation (NLG) systems. Nevertheless, the evaluation guideline, as a pivotal element ensuring reliable and reproducible human assessment, has received limited attention.Our investigation revealed that only 29.84% of recent papers involving human evaluation at top conferences release their evaluation guidelines, with vulnerabilities identified in 77.09% of these guidelines. Unreliable evaluation guidelines can yield inaccurate assessment outcomes, potentially impeding the advancement of NLG in the right direction. To address these challenges, we take an initial step towards reliable evaluation guidelines and propose the first human evaluation guideline dataset by collecting annotations of guidelines extracted from existing papers as well as generated via Large Language Models (LLMs). We then introduce a taxonomy of eight vulnerabilities and formulate a principle for composing evaluation guidelines. Furthermore, a method for detecting guideline vulnerabilities has been explored using LLMs, and we offer a set of recommendations to enhance reliability in human evaluation. The annotated human evaluation guideline dataset and code for the vulnerability detection method are publicly available online.

Anya Belz, Craig Thomson, Ehud Reiter et al.

We report our efforts in identifying a set of previous human evaluations in NLP that would be suitable for a coordinated study examining what makes human evaluations in NLP more/less reproducible. We present our results and findings, which include that just 13\% of papers had (i) sufficiently low barriers to reproduction, and (ii) enough obtainable information, to be considered for reproduction, and that all but one of the experiments we selected for reproduction was discovered to have flaws that made the meaningfulness of conducting a reproduction questionable. As a result, we had to change our coordinated study design from a reproduce approach to a standardise-then-reproduce-twice approach. Our overall (negative) finding that the great majority of human evaluations in NLP is not repeatable and/or not reproducible and/or too flawed to justify reproduction, paints a dire picture, but presents an opportunity for a rethink about how to design and report human evaluations in NLP.

Zhun Fan, Zhaojun Wang, Wenji Li et al.

In dealing with constrained multi-objective optimization problems (CMOPs), a key issue of multi-objective evolutionary algorithms (MOEAs) is to balance the convergence and diversity of working populations.

Zhun Fan, Wenji Li, Zhaojun Wang et al.

This paper proposes a push and pull search method in the framework of differential evolution (PPS-DE) to solve constrained single-objective optimization problems (CSOPs). More specifically, two sub-populations, including the top and bottom sub-populations, are collaborated with each other to search global optimal solutions efficiently. The top sub-population adopts the pull and pull search (PPS) mechanism to deal with constraints, while the bottom sub-population use the superiority of feasible solutions (SF) technique to deal with constraints. In the top sub-population, the search process is divided into two different stages --- push and pull stages.An adaptive DE variant with three trial vector generation strategies is employed in the proposed PPS-DE. In the top sub-population, all the three trial vector generation strategies are used to generate offsprings, just like in CoDE. In the bottom sub-population, a strategy adaptation, in which the trial vector generation strategies are periodically self-adapted by learning from their experiences in generating promising solutions in the top sub-population, is used to choose a suitable trial vector generation strategy to generate one offspring. Furthermore, a parameter adaptation strategy from LSHADE44 is employed in both sup-populations to generate scale factor $F$ and crossover rate $CR$ for each trial vector generation strategy. Twenty-eight CSOPs with 10-, 30-, and 50-dimensional decision variables provided in the CEC2018 competition on real parameter single objective optimization are optimized by the proposed PPS-DE. The experimental results demonstrate that the proposed PPS-DE has the best performance compared with the other seven state-of-the-art algorithms, including AGA-PPS, LSHADE44, LSHADE44+IDE, UDE, IUDE, $ε$MAg-ES and C$^2$oDE.