Austria

This work aims to understand how scaling improves language models, specifically in terms of training dynamics. We find that language models undergo loss deceleration early in training—an abrupt slowdown in the rate of loss improvement, resulting in piecewise linear behaviour of the loss curve in log-log space. Scaling up the model mitigates this transition by (1) decreasing the loss at which deceleration occurs, and (2) improving the log-log rate of loss improvement after deceleration. We attribute loss deceleration to a type of degenerate training dynamics we term zero-sum learning (ZSL). In ZSL, per-example gradients become systematically opposed, leading to destructive interference in per-example changes in loss. As a result, improving loss on one subset of examples degrades it on another, bottlenecking overall progress. Loss deceleration and ZSL provide new insights into the training dynamics underlying language model scaling laws, and could potentially be targeted directly to improve language models independent of scale. We make our code and artefacts available at: https://github.com/mirandrom/zsl

ACL 2025

Understanding Language Model Scaling Laws in Terms of Training Dynamics via Loss Deceleration and Zero-Sum Learning

scaling laws

training dynamics

language models

## 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.

technical paper

### 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)

Data augmentation is an essential technique in natural language processing (NLP) for enriching training datasets by generating diverse samples. This process is crucial for improving the robustness and generalization capabilities of NLP models. However, a significant challenge remains: Insufficient Attention to Sample Distribution Diversity. Most existing methods focus on increasing the sample numbers while neglecting the sample distribution diversity, which can lead to model overfitting. In response, we explore data augmentation's impact on dataset diversity and propose a Diversity-oriented data Augmentation framework (DoAug). Specifically, we utilize a diversity-oriented fine-tuning approach to train a large language model (LLM) as a diverse paraphraser, which is capable of augmenting textual datasets by generating diversified paraphrases. Then, we apply the LLM paraphraser to a selected coreset of highly informative samples and integrate the paraphrases with the original data to create a more diverse augmented dataset. Finally, we conduct extensive experiments on 12 real-world textual datasets. The results show that our fine-tuned LLM augmenter improves diversity while preserving label consistency, thereby enhancing the robustness and performance of downstream tasks. Specifically, it achieves an average performance gain of 10.52%, surpassing the runner-up baseline with more than three percentage points.

Diversity-oriented Data Augmentation with Large Language Models

Precise control over language model generation is vital for ensuring both safety and reliability. 
Although prompt engineering and steering are commonly used to intervene in model behaviors, the vast number of parameters in models often results in highly intertwined internal representations. 
This interdependency can limit control precision and sometimes lead to unintended side effects. Recent research has explored the use of sparse autoencoders (SAE) to disentangle knowledge in high-dimensional spaces for steering.
However, these applications have been limited to toy tasks owing to the nontrivial issue of locating ``atomic knowledge components''. 
In this paper, we propose Steering Target Atoms (STA), a novel method that isolates and manipulates disentangled knowledge components to enhance safety. Comprehensive experiments demonstrate the effectiveness of our approach. 
Further analysis reveals that steering exhibits superior robustness and flexibility, particularly in adversarial scenarios. 
We also apply the steering strategy to the large reasoning model, confirming its effectiveness in precise reasoning control.

Beyond Prompt Engineering: Robust Behavior Control in LLMs via Steering Target Atoms

While it is commonly accepted that maintaining common ground plays a role in conversational success, little prior research exists connecting conversational grounding to success in task-oriented conversations. We study failures of grounding in the Ubuntu IRC dataset, where participants use text-only communication to resolve technical issues. We find that disruptions in conversational flow often stem from a misalignment in common ground, driven by a divergence in beliefs and assumptions held by participants. These disruptions, which we call conversational friction, significantly correlate with task success. While LLMs can identify overt cases of conversational friction, they struggle with subtler and more context-dependent instances that require pragmatic or domain-specific reasoning.

Understanding Common Ground Misalignment in Goal-Oriented Dialog: A Case-Study with Ubuntu Chat Logs

Large Language Models (LLMs) have demonstrated impressive capabilities in generating coherent text but often struggle with grounding language and strategic dialogue. To address this gap, we focus on journalistic interviews, a domain rich in grounding communication and abundant in data. We curate a dataset of 40,000 two-person informational interviews from NPR and CNN, and reveal that LLMs are significantly less likely than human interviewers to use acknowledgements and to pivot to higher-level questions. Realizing that a fundamental deficit exists in multi-turn planning and strategic thinking, we develop a realistic simulated environment, incorporating source personas and persuasive elements, in order to facilitate the development of agents with longer-horizon rewards. Our experiments show that while source LLMs mimic human behavior in information sharing, interviewer LLMs struggle with recognizing when questions are answered and engaging persuasively, leading to suboptimal information extraction across model size and capability. These findings underscore the need for enhancing LLMs' strategic dialogue capabilities.

NewsInterview: a Dataset and a Playground to Evaluate LLMs' Grounding Gap via Informational Interviews

Electronic Discovery (eDiscovery) requires identifying relevant documents from vast collections for legal production requests. While artificial intelligence (AI) and natural language processing (NLP) have improved document review efficiency, current methods still struggle with legal entities, citations, and complex legal artifacts. To address these challenges, we introduce DISCOvery Graph (DISCOG), an emerging system that integrates knowledge graphs for enhanced document ranking and classification, augmented by LLM-driven reasoning. DISCOG outperforms strong baselines in F1-score, precision, and recall across both balanced and imbalanced datasets. In real-world deployments, it has reduced litigation-related document review costs by approximately 98%, demonstrating significant business impact.

Learning from Litigation: Graphs for Retrieval and Reasoning in eDiscovery

Recent advancements in latent diffusion models (LDMs) have markedly enhanced text-to-audio generation, yet their iterative sampling processes impose substantial computational demands, limiting practical deployment. While recent methods utilizing consistency-based distillation aim to achieve few-step or single-step inference, their one-step performance is constrained by curved trajectories, preventing them from surpassing traditional diffusion models. In this work, we introduce FlashAudio with rectified flows to learn straight flow for fast simulation. To alleviate the inefficient timesteps allocation and suboptimal distribution of noise, FlashAudio optimizes the time distribution of rectified flow with Bifocal Samplers and proposes immiscible flow to minimize the total distance of data-noise pairs in a batch vias assignment. Furthermore, to address the amplified accumulation error caused by the classifier-free guidance (CFG), we propose Anchored Optimization, which refines the guidance scale by anchoring it to a reference trajectory. Experimental results on text-to-audio generation demonstrate that FlashAudio's one-step generation performance surpasses the diffusion-based models with hundreds of sampling steps on audio quality and enables a sampling speed of 400x faster than real-time on a single NVIDIA 4090Ti GPU. Code will be available at~\url{https://github.com/liuhuadai/FlashAudio}. Audio Samples are available at https://FlashAudio-TTA.github.io/.

FlashAudio: Rectified Flow for Fast and High-Fidelity Text-to-Audio Generation

Bragging, as a pervasive social-linguistic phenomenon, reflects complex human interaction patterns. However, the understanding and generation of appropriate bragging behavior in large language models (LLMs) remains underexplored. In this paper, we propose a comprehensive study that combines analytical and controllable approaches to examine bragging in LLMs. We design three tasks, \textit{bragging recognition, bragging explanation, and bragging generation}, along with novel evaluation metrics to assess the models' ability to identify bragging intent, social appropriateness, and account for context sensitivity. Our analysis reveals the challenges of bragging in the social context, such as recognizing bragging and responding appropriately with bragging in conversation. This work provides new insights into how LLMs process bragging and highlights the need for more research on generating contextually appropriate behavior in LLMs.

It’s Not Bragging If You Can Back It Up: Can LLMs Understand Braggings?

With the exponential growth of research facilitated by modern technology and improved accessibility, scientific discoveries have become increasingly fragmented within and across fields. This makes it challenging to assess the significance, novelty, incremental findings, and equivalent ideas between related works, particularly those from different research communities. Large language models (LLMs) have recently demonstrated strong quantitative and qualitative reasoning abilities, and multi-agent LLM debates have shown promise in handling complex reasoning tasks by exploring diverse perspectives and reasoning paths. Inspired by this, we introduce Tree-of-Debate (ToD), a framework which converts scientific papers into LLM personas that debate their respective novelties. To emphasize structured, critical reasoning rather than focusing solely on outcomes, ToD dynamically constructs a debate tree, enabling fine-grained analysis of independent novelty arguments within scholarly articles. Through experiments on scientific literature across various domains, evaluated by expert researchers, we demonstrate that ToD generates informative arguments, effectively contrasts papers, and supports researchers in their literature review.

Tree-of-Debate: Multi-Persona Debate Trees Elicit Critical Thinking for Scientific Comparative Analysis

Keyboard-based interaction may not accommodate various needs, especially for individuals with disabilities. While inertial sensor-based writing recognition is promising due to the sensors' small size, wearability, and low cost, accurate recognition in the Chinese context is hampered by the difficulty of collecting extensive inertial signal samples for the vast number of characters. Therefore, we design a Chinese Inertial GAN (CI-GAN) containing Chinese glyph encoding (CGE), forced optimal transport (FOT), and semantic relevance alignment (SRA) to acquire unlimited high-quality training samples. Unlike existing vectorization methods focusing on the meaning of Chinese characters, CGE represents shape and stroke features, providing glyph guidance for writing signal generation. FOT establishes a triple-consistency constraint between the input prompt, output signal features, and real signal features, ensuring the authenticity and semantic accuracy of the generated signals. SRA aligns semantic relationships between multiple outputs and their input prompts, ensuring that similar inputs correspond to similar outputs (and vice versa), alleviating model hallucination. The three modules guide the generator while also interacting with each other, forming a coupled system. By utilizing the massive training samples provided by CI-GAN, the performance of six widely used classifiers is improved from 6.7\% to 98.4\%, indicating that CI-GAN constructs a flexible and efficient data platform for Chinese inertial writing recognition. Furthermore, we release the first Chinese inertial writing dataset on GitHub.

Chinese Inertial GAN for Handwriting Signal Generation and Recognition

In historical linguistics, the affiliation of languages to a common language family is traditionally carried out using a complex workflow that relies on manually comparing individual languages. Large-scale standardized collections of multilingual wordlists and grammatical language structures might help to improve this and open new avenues for developing automated language affiliation workflows. Here, we present neural network models that use lexical and grammatical data from a worldwide sample of more than 1,200 languages with known affiliations to classify individual languages into families. In line with the traditional assumption of most linguists, our results show that models trained on lexical data alone outperform models solely based on grammatical data, whereas combining both types of data yields even better performance. In additional experiments, we show how our models can identify long-ranging relations between entire subgroups, how they can be employed to investigate potential relatives of linguistic isolates, and how they can help us to obtain first hints on the affiliation of so far unaffiliated languages. We conclude that models for automated language affiliation trained on lexical and grammatical data provide comparative linguists with a valuable tool for evaluating hypotheses about deep and unknown language relations.

Downloads

Next from ACL 2025

Diversity-oriented Data Augmentation with Large Language Models

Similar lecture

Refinement Matters: Textual Description Needs to be Refined for Zero-shot Learning