Austria

Recent advancements in large language models (LLMs) have significantly enhanced their ability to understand both natural language and code, driving their use in tasks like natural language-to-code (NL2Code) and code summarisation. However, LLMs are prone to hallucination—outputs that stray from intended meanings. Detecting hallucinations in code summarisation is especially difficult due to the complex interplay between programming and natural languages. We introduce a first-of-its-kind dataset, CodeSumEval, with $\sim$10K samples, curated specifically for hallucination detection in code summarisation. We further propose a novel Entity Tracing Framework (ETF) that a) utilises static program analysis to identify code entities from the program and b) uses LLMs to map and verify these entities and their intents within generated code summaries. Our experimental analysis demonstrates the framework’s effectiveness, leading to a 73\% F1 score. The proposed approach provides a method for detecting hallucinations by tracing entities from the summary to the code, allowing us to evaluate summary accuracy and localise the error within the summary.

ACL 2025

ETF: An Entity Tracing Framework for Hallucination Detection in Code Summaries

code understanding

hallucination

code summarization

## Welcome to EMNLP 2022!
I am delighted to welcome you to EMNLP 2022! I believe this conference has been complicated beyond any precedent. Over the past year, it’s been thrilling to see the organization team approach each new puzzle with creativity and enthusiasm. We hope that those participating in Abu Dhabi as well as those joining remotely will leave the conference feeling newly inspired by the program and newly connected to our ever-growing community. Following EMNLP 2021 and major NLP conferences since, EMNLP 2022 is “hybrid,” serving both virtual and in-person participants.

Our key innovations for EMNLP 2022 include:

* EMNLP 2022 is “hybrid” in a second sense, as well: we allowed both direct and rolling review paper submissions, building on the pilot experiment of EMNLP 2021, which considered a small number of ARR submissions. 
* Familiar from NAACL but new to EMNLP, we’ve added an industry track.
* During the conference, “portals” will link virtual poster sessions to in-person conference participants during poster sessions each day.
* The first ACL-family conference in the United Arab Emirates.

 *Message from Noah A. Smith, University of Washington and Allen Institute for AI, Seattle, Washington, USA* 
***EMNLP 2022 General Chair***
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/9eec7d4a287ee18c278b08229290aa83.png)](https://drive.google.com/file/d/1OlPv6QBeo62VVTughj2jkiLeyHd1WnUt/view)
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/a3db7a768409f05192210d98601edb25.png)](https://emnlp2022.rocket.chat/)

To access this site you need to register. Please register [here](https://2022.emnlp.org/registration/).

Register here

EMNLP 2022

Welcome!
EMNLP 2022 will take place in Abu Dhabi from December 7th to December 11th, 2022. And it will be held in hybrid mode, both online and offline.

poster

### Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Message from the General Chair: 
*It is my great pleasure and honor to welcome you to the 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025), held in beautiful Vienna, Austria, from July 27 to August 1, 2025. ACL2025continues our field’s tradition of excellence in scholarship, innovation, and inclusivity, and I am deeply grateful to the many volunteers who have worked tirelessly to bring this event to life.* 
[Read more](https://drive.google.com/file/d/1GI_hvOpjswAuYdUTromfeDiPpCcqidwg/view?usp=sharing)

To access this event page, you need to log in with the **email address you registered with**. Access credentials will be sent to your email from Underline - subject line "Welcome to ACL 2025". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you need to log in with the **email address you registered with**. 

Welcome to The 63rd Annual Meeting of the Association for Computational Linguistics (ACL 2025)

Reinforcement Learning (RL) algorithms for safety alignment of Large Language Models (LLMs), such as Direct Preference Optimization (DPO), encounter the challenge of distribution shift. 
Current approaches typically address this issue through online sampling from the target policy, which requires significant computational resources.
In this paper, we hypothesize that during off-policy training, while the ranking order of output generated by policy changes, their overall distribution remains relatively stable.
This stability allows the conversion of the sampling process from the target policy into a computationally
efficient re-ranking of preference data.
Building on this hypothesis, we propose a new framework that leverages the model's intrinsic safety judgment capability to extract reward signals, which are then used to calculate label confidence for preference reordering. 
Extensive experiments and theoretical analysis demonstrate that the proposed method effectively addresses the distribution shift issue, 
remarkably enhancing the safety performance while avoiding about 300x computational overheads.

Representation-based Reward Modeling for Efficient Safety Alignment of Large Language Model

We introduce a professionally translated extension of the TruthfulQA benchmark designed to evaluate truthfulness in Basque, Catalan, Galician, and Spanish. Truthfulness evaluations of large language models (LLMs) have primarily been focused on English. However, the ability of LLMs to maintain truthfulness across languages remains under-explored. Our study evaluates 12 state-of-the-art open LLMs, comparing base and instruction-tuned models using human evaluation, multiple-choice metrics, and LLM-as-a-Judge scoring. Our findings reveal that, while LLMs perform best in English and worst in Basque (the lowest-resourced language), overall truthfulness discrepancies across languages are smaller than anticipated. Furthermore, we show that LLM-as-a-Judge correlates more closely with human judgments than multiple-choice metrics, and that informativeness plays a critical role in truthfulness assessment. Our results also indicate that machine translation provides a viable approach for extending truthfulness benchmarks to additional languages, offering a scalable alternative to professional translation. Finally, we observe that universal knowledge questions are better handled across languages than context- and time-dependent ones, highlighting the need for truthfulness evaluations that account for cultural and temporal variability. Datasets, models and code are publicly available under open licenses.

Truth Knows No Language: Evaluating Truthfulness Beyond English

Misinformation can be countered with fact-checking, but the process is costly and slow. Identifying checkworthy claims is the first step, where automation can help scale fact-checkers' efforts. However, detection methods struggle with content that is (1) multimodal, (2) from diverse domains, and (3) synthetic. We introduce HintsOfTruth, a public dataset for multimodal checkworthiness detection with 27K real-world and synthetic image/claim pairs. The mix of real and synthetic data makes this dataset unique and ideal for benchmarking detection methods. We compare fine-tuned and prompted Large Language Models (LLMs). We find that well-configured lightweight text-based encoders perform comparably to multimodal models but the former only focus on identifying non-claim-like content. Multimodal LLMs can be more accurate but come at a significant computational cost, making them impractical for large-scale applications. When faced with synthetic data, multimodal models perform more robustly.

HintsOfTruth: A Multimodal Checkworthiness Detection Dataset with Real and Synthetic Claims

Multimodal coreference resolution (MCR) aims to identify mentions referring to the same entity across different modalities, such as text and visuals, and is essential for understanding multimodal content. In the era of rapidly growing multimodal content and social media, MCR is particularly crucial for interpreting user interactions and bridging text-visual references to improve communication and personalization. However, MCR research for real-world dialogues remains unexplored due to the lack of sufficient data resources. To address this gap, we introduce TikTalkCoref, the first Chinese multimodal coreference dataset for social media in real-world scenarios, derived from the popular Douyin short-video platform. This dataset pairs short videos with corresponding textual dialogues from user comments and includes manually annotated coreference clusters for both person mentions in the text and the coreferential person head regions in the corresponding video frames. We also present an effective benchmark approach for MCR, focusing on the celebrity domain, and conduct extensive experiments on our dataset, providing reliable benchmark results for this newly constructed dataset. We release the TikTalkCoref dataset to facilitate future research on MCR for real-world social media dialogues at https://github.com/lxystaruni/TikTalkCoref.

Multimodal Coreference Resolution for Chinese Social Media Dialogues: Dataset and Benchmark Approach

Text-based reinforcement-learning agents improve their policies by interacting with their environments to collect more training data. However, these self-collected data inevitably contain intermediate failed actions caused by attempting physically infeasible behaviors and/or hallucinations. Directly learning a policy from such trajectories can reinforce incorrect behaviors and reduce task success rates. In this paper, we propose a failed action-aware objective that suppresses the negative impact of failed actions during training by assigning zero return based on textual feedback. Building on this objective, we introduce a perturbation method that leverages unsuccessful trajectories to construct new successful ones that share the same goal. This allows agents to benefit from diverse experiences without further interaction with the environment. Experiments in ALFWorld and ScienceWorld demonstrate that our method significantly outperforms strong baselines and generalizes across environments. Code is available at https://github.com/riken-grp/text-agent.

Teaching Text Agents to Learn Sequential Decision Making from Failure

A central question in multilingual language modeling is whether large language models (LLMs) develop a universal concept representation, disentangled from specific languages. In this paper, we address this question by analyzing latent representations (latents) during a word-translation task in transformer-based LLMs. We strategically extract latents from a source translation prompt and insert them into the forward pass on a target translation prompt. By doing so, we find that the output language is encoded in the latent at an earlier layer than the concept to be translated. Building on this insight, we conduct two key experiments. First, we demonstrate that we can change the concept without changing the language and vice versa through activation patching alone. Second, we show that patching with the mean representation of a concept across different languages does not affect the models' ability to translate it, but instead improves it. Finally, we generalize to multi-token generation and demonstrate that the model can generate natural language description of those mean representations. Our results provide evidence for the existence of language-agnostic concept representations within the investigated models.

Separating Tongue from Thought: Activation Patching Reveals Language-Agnostic Concept Representations in Transformers

Enhancing the reasoning capabilities of language models (LMs) remains a key challenge, especially for tasks that require complex, multi-step decision-making where existing Chain-of-Thought (CoT) approaches struggle with consistency and verification. In this paper, we propose a novel reasoning framework, referred to as Structure-aware Planning with an Accurate World Model (SWAP), that integrates structured knowledge representation with learned planning. Unlike prior methods that rely purely on natural language reasoning, SWAP leverages entailment graphs to encode structured dependencies and enable symbolic verification of intermediate steps. To systematically construct and update the graph, SWAP employs a policy model to propose candidate expansions and a world model to predict structural updates. To improve accuracy, the world model generates multiple alternative updates, and a discriminator re-ranks them based on plausibility. To encourage diverse exploration, we introduce Diversity-based Modelling (DM), which samples candidates from the remaining probability mass after removing previously sampled candidates from the original policy distribution. Additionally, SWAP improves the discrimination accuracy through Contrastive Ranking (CR), which directly compares candidates within prompts and incorporates meta-knowledge to improve ranking quality. We evaluate SWAP across diverse reasoning-intensive benchmarks including math reasoning, logical reasoning, and coding tasks. Extensive experiments demonstrate that SWAP significantly improves upon the base models and consistently outperforms existing reasoning methods.

Deliberate Reasoning for Language Models as Structure-aware Planning with Accurate World Model

In this work, we introduce the PKU-SafeRLHF dataset, designed to promote research on safety alignment in large language models (LLMs). As a sibling project to SafeRLHF and BeaverTails, we separate annotations of helpfulness and harmlessness for question-answering pairs, providing distinct perspectives on these coupled attributes. Overall, we provide 44.6k refined prompts and 265k question-answer pairs with safety meta-labels for 19 harm categories and three severity levels ranging from minor to severe, with answers generated by Llama-family models. Based on this, we collected 166.8k preference data, including dual-preference (helpfulness and harmlessness decoupled) and single-preference data (trade-off the helpfulness and harmlessness from scratch), respectively. Using the large-scale annotation data, we further train severity-sensitive moderation for the risk control of LLMs and safety-centric RLHF algorithms for the safety alignment of LLMs. We believe this dataset will be a valuable resource for the community, aiding in the safe deployment of LLMs.

BeaverTails v2: Towards Multi-Level Safety Alignment for LLMs with Human Preference

The choice of tokenizer can profoundly impact language model performance, yet accessible and reliable evaluations of tokenizer quality remain an open challenge. Inspired by scaling consistency, we show that smaller models can accurately predict significant differences in tokenizer impact on larger models at a fraction of the compute cost. By systematically evaluating both English-centric and multilingual tokenizers, we find that tokenizer choice has negligible effects on tasks in English but results in consistent performance differences in multilingual settings. We propose new intrinsic tokenizer metrics inspired by Zipf's law that correlate more strongly with downstream performance than text compression when modeling unseen languages. By combining several metrics to capture multiple aspects of tokenizer behavior, we develop a reliable framework for intrinsic tokenizer evaluations. Our work offers a more efficient path to informed tokenizer selection in future language model development.

Beyond Text Compression: Evaluating Tokenizers Across Scales

Continued pretraining (CPT) is a popular approach to adapt existing large language models (LLMs) to new languages. When doing so, it is common practice to include a portion of English data in the mixture, but its role has not been carefully studied to date. In this work, we show that including English does not impact validation perplexity, yet it is critical for the emergence of downstream capabilities in the target language. We introduce a language-agnostic benchmark for in-context learning (ICL), which reveals catastrophic forgetting early on CPT when English is not included. This in turn damages the ability of the model to generalize to downstream prompts as measured by perplexity, even if it does not manifest in terms of accuracy until later in training, and can be tied to a big shift in the model parameters. Based on these insights, we introduce curriculum learning and exponential moving average (EMA) of weights as effective alternatives to mitigate the need for English. All in all, our work sheds light into the dynamics by which emergent abilities arise when doing CPT for language adaptation, and can serve as a foundation to design more effective methods in the future.

Downloads

Next from ACL 2025

Representation-based Reward Modeling for Efficient Safety Alignment of Large Language Model

Similar lecture

Improved Evaluation of Automatic Source Code Summarisation