United States

This paper narrows the performance gap between small, specialized models and significantly larger general-purpose models through domain adaptation via continual pre-training and merging. We address the scarcity of specialized non-English data by constructing a high-quality German medical corpus (FineMed-de) from FineWeb2. This corpus is used to continually pre-train and merge three well-known LLMs (ranging from 7B to 24B parameters), creating the DeFineMed model family. A comprehensive evaluation confirms that specialization dramatically enhances 7B model performance on German medical benchmarks. Furthermore, the pairwise win-rate analysis of the Qwen2.5-based models demonstrates an approximately 3.5-fold increase in the win-rate against the much larger Mistral-Small-24B-Instruct through domain adaptation. This evidence positions specialized 7B models as a competitive, resource-efficient solution for complex medical instruction-following tasks. While model merging successfully restores instruction-following abilities, a subsequent failure mode analysis reveals inherent trade-offs, including the introduction of language mixing and increased verbosity, highlighting the need for more targeted fine-tuning in future work. This research provides a robust, compliant methodology for developing specialized LLMs, serving as the foundation for practical use in German-speaking healthcare contexts.

ACL 2026

Can Continual Pretraining Bridge the Performance Gap between General-purpose and Specialized Language Models in the Medical Domain?

## Welcome!
"I am excited to welcome you to this year’s edition of the Conference on Empirical Methods in Natural Language Processing! Importantly, it marks the 30th edition of EMNLP. With over 8,000 submissions, more than 3,000 accepted papers, and thousands of attendees, we have come a long way from that first
workshop, which had 14 accepted papers. As the field looks ahead, Suzhou is the fitting location for celebrating this milestone: rooted in a long literary tradition, yet modern and forward-looking, and home to a large share of the NLP community." 

*Message from the General Chair, Dirk Hovy*

EMNLP 2025

Celebrate 30 Years of EMNLP! 
EMNLP 2025 will be held in Suzhou, China from November 5th to November 9th, 2025.

poster

It is my great pleasure and honor to welcome you to the 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026), held in San Diego, California, USA, July 2–7, 2026. ACL 2026 continues our fields 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.

It is exciting to see the many developments in the field reflected at our conference, such as the advancement of large language models (born out of our community) into many application directions. This is not without challenges, partly due to unsolved problems, and partly due to misuse of the technology. All this is reflected well at ACL 2026. The conference also reflects that the field is not only about advancing technology, but also the science of fundamentally understanding language and how it is use by many different speaker communities. 
[Full Message from General Chair Philipp Koehn](https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing) 
<html><button style="display: inline-flex; align-items: center; justify-content: center; white-space: nowrap; border-radius: 9999px; font-weight: bold; background: #7c3aed; color: white; font-family: 'Space Grotesk', sans-serif; height: 40px; font-size: 16px; padding: 0 20px; border: none; cursor: pointer" onclick="window.open('https://drive.google.com/file/d/1QiJFOmMjvDgsDo3i1XzlQnImxe3RU_Cy/view?usp=sharing')">Conference Handbook</button></html>

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 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://2026.aclweb.org/registration/) first.

**Online Registration Form**: https://acl.swoogo.com/acl2026

Registration Required

The 64th Annual Meeting of the Association for Computational Linguistics (ACL 2026) will take place in San Diego, California, from July 2 - 7, 2026.

Reasoning over long contexts remains a major challenge for language models, particularly when solving tasks that require integrating multiple facts in sequence or generalizing to new distributions. We argue that this difficulty stems from a lack of structural inductive bias. Recently, alternative frameworks have been proposed to explicitly encode contexts as ordered memory and perform iterative retrieval to construct reasoning chains. Despite the promising results shown in prior arts, they are still heavily reliant on intermediate chain supervision and fall short in showing emergent reasoning generalization in the presence of hard distractions in reasoning-in-a-haystack tasks. Furthermore, we discover that as the amount of distractions increases, traditional episodic memory reads suffer from ill-conditioning problems, which lead to inaccurate context retrievals. In this work, we formalize the motivation for necessary inductive bias in reasoning-in-a-Haystack tasks, propose inference-time memory update procedures mimicking the "identify and remove unnecessary and unrelated details" in *constructively responsive reading*, introduce staged training inspired by human conceptual understanding, and finally demonstrate the possibilities and limits of such framework in the weakly supervised scenario.

ImReasoner: Improving Memory-based Language Models for Reasoning-in-a-Haystack Tasks

Translating software written in C to Rust has significant benefits in improving memory safety. However, manual translation is cumbersome, error-prone, and often produces unidiomatic code. Large language models (LLMs) have demonstrated promise in producing idiomatic translations, but offer no correctness guarantees. We propose SACTOR, an LLM-driven C-to-Rust translation tool that employs a two-step process: an initial "unidiomatic" translation to preserve semantics, followed by an "idiomatic" refinement to align with Rust standards. SACTOR leverages static analysis of the C source to handle pointer semantics and dependency resolution. To validate correctness of our function-wise incremental translation that mixes C and Rust, we use end-to-end testing via the foreign function interface. We evaluate SACTOR on $200$ programs from two public datasets and on two more complex scenarios (a 50-sample subset of CRust-Bench and the libogg library), comparing multiple LLMs. Across datasets, SACTOR delivers high end-to-end correctness and produces safe, idiomatic Rust with up to 7$\times$ fewer Clippy warnings; On CRust-Bench, SACTOR achieves an average (across samples) of 85% unidiomatic and 52% idiomatic success, and on libogg it attains full unidiomatic and up to 78% idiomatic coverage on GPT-5.

SACTOR: LLM-Driven Correct and Idiomatic C to Rust Translation with Static Analysis and FFI-Based Verification

Multi-modal large language models (MLLMs) have emerged as powerful tools for analyzing Internet-scale image data, offering significant benefits but also raising critical safety and societal concerns. In particular, these models may be misused to extract sensitive information from personal images, such as identifying individuals or revealing locations. In this work, we propose ImageProtector, a method designed to protect images from unauthorized analysis by MLLMs. Before an image is shared online, ImageProtector embeds a carefully crafted, nearly imperceptible perturbation that acts as a visual prompt injection attack on MLLMs. Consequently, when a malicious actor downloads and queries a protected image, the MLLM is consistently misled into generating a refusal response such as "I’m sorry, I can’t help with that request." We empirically demonstrate the effectiveness of ImageProtector across six MLLMs and four datasets. Additionally, we evaluate three potential countermeasures, Gaussian noise, DiffPure, and adversarial training, and show that while they partially mitigate the impact of ImageProtector, they simultaneously degrade model accuracy and/or efficiency.

Leave My Images Alone: Preventing Multi-Modal Large Language Models from Analyzing Images via Visual Prompt Injection

Reranking algorithms have made progress in improving document retrieval quality by efficiently aggregating relevance judgments generated by large language models (LLMs). However, identifying relevant documents for queries that require in-depth reasoning remains a major challenge. Reasoning‐intensive queries often exhibit multifaceted information needs and nuanced interpretations, rendering document relevance inherently context dependent and often noisy. To address this, we propose contextual relevance, which we define as the probability that a document is relevant to a given query, marginalized over the distribution of different reranking contexts it may appear in (i.e., the set of candidate documents it is ranked alongside and the order in which the documents are presented to a reranking model). While prior works have studied methods to mitigate the positional bias LLMs exhibit by accounting for the ordering of documents, we empirically show that batch composition also materially affects relevance judgments. To efficiently estimate contextual relevance, we propose TS-SetRank, a sampling-based, uncertainty-aware reranking algorithm. Empirically, TS-SetRank improves nDCG@10 over retrieval and reranking baselines by 15–25% on BRIGHT and 6–21% on BEIR, highlighting the importance of modeling relevance as context-dependent.

Contextual Relevance and Adaptive Sampling for LLM-Based Document Reranking

While self-consistency is known to improve performance on symbolic reasoning, its effect on the recall of encyclopedic knowledge is unclear due to a lack of targeted evaluation grounds.
To address this, we establish such a knowledge recall split for the popular MMLU benchmark by applying a data-driven heuristic from prior work.
We validate this split by showing that the performance patterns on the symbolic reasoning and knowledge recall subsets mirror those of GSM8K and MedMCQA, respectively.
Using this solid ground, we find that self-consistency consistently improves performance across both symbolic reasoning and knowledge recall, even though its underlying CoT prompting is primarily effective for symbolic reasoning.
As a result, we achieve an 89\% accuracy on MMLU, the best performance to date with the use of GPT-4o.

Does Self-Consistency Improve the Recall of Encyclopedic Knowledge?

Reinforcement learning with verifiable rewards (RLVR) has recently advanced the reasoning capabilities of large language models (LLMs). We investigate RLVR from a sample-centric perspective and introduce **LPPO** (Learning-Progress and Prefix-guided Optimization), a framework of progressive optimization techniques. Our work addresses a critical question: how to best leverage a small set of trusted, high-quality demonstrations, rather than simply scaling up data volume. First, motivated by how hints aid human problem-solving, we propose **prefix-guided sampling**, an online data augmentation method that incorporates partial solution prefixes from expert demonstrations to guide the policy, particularly for challenging instances. Second, inspired by how humans focus on important questions aligned with their current capabilities, we introduce **learning-progress weighting**, a dynamic strategy that adjusts each training sample's influence based on model progression. We estimate sample-level learning progress via an exponential moving average of per-sample pass rates, promoting samples that foster learning and de-emphasizing stagnant ones. Experiments on mathematical-reasoning benchmarks demonstrate that our methods outperform strong baselines, yielding faster convergence and a higher performance ceiling, with these gains proving robust across diverse model architectures, scales, and reinforcement learning optimizers.

From Data-Centric to Sample-Centric: Enhancing LLM Reasoning via Progressive Optimization

Robust deployment of large multimodal models (LMMs) in real-world scenarios requires access to external knowledge sources, given the complexity and dynamic nature of real-world information. Existing approaches such as retrieval-augmented generation (RAG) and prompt engineered search agents rely on rigid pipelines, often leading to inefficient or excessive search behaviors. We present MMSearch-R1, the first end-to-end reinforcement learning framework that enables LMMs to perform on-demand, multi-turn search in real-world Internet environments. Our framework integrates both image and text search tools, allowing the model to reason about when and how to invoke them guided by an outcome-based reward with a search penalty. To support training, We collect a multimodal search VQA dataset through a semi-automated pipeline that covers diverse visual and textual knowledge needs and curate a search-balanced subset with both search-required and search-free samples, which proves essential for shaping efficient and on-demand search behavior. Extensive experiments on knowledge-intensive and info-seeking VQA tasks show that our model not only outperforms traditional RAG-based baselines of the same model size, but also matches the performance of a larger RAG-based model while reducing search calls by over 30%. We further analyze key empirical findings to offer actionable insights for advancing research in multimodal search.

MMSearch-R1: Incentivizing LMMs to Search

Pretraining massively multilingual Large Language Models (LLMs) for many languages at once is challenging due to limited model capacity, scarce high-quality data, and compute constraints. Moreover, the lack of language coverage in the tokenizer makes it harder to address the gap for new languages purely at the post-training stage. In this work, we study what relatively cheap interventions early on in training improve *language plasticity*, or adaptation capabilities of the model post-training to new languages. We focus on tokenizer design and propose using a *universal* tokenizer that is trained for more languages than the primary pretraining languages to enable efficient adaptation in expanding language coverage after pretraining. Our systematic experiments across diverse groups of languages and different training strategies show that a universal tokenizer enables significantly higher language adaptation, with up to 20.2\% increase in win rates compared to tokenizers specific to pretraining languages. Furthermore, a universal tokenizer also leads to better plasticity towards languages that are completely unseen in the tokenizer and pretraining, by up to 5\% win rate gain. We achieve this adaptation to an expanded set of languages with minimal compromise in performance on the majority of languages included in pretraining.

One Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual Tokenizers

Users frequently express their beliefs to large language models (LLMs). In some situations, the LLM should accept these contextual beliefs as true. In others, they should stick to their prior knowledge. Notably, users' expressions of belief (EoBs) can take linguistically diverse forms---using presuppositions, evidential and certainty markers, or varied tones---each of which may have a different persuasiveness over the LLMs. We introduce a typology to systematically evaluate how different EoBs affect whether models follow context versus prior knowledge. The typology is grounded in four linguistically motivated dimensions: form, evidentiality, epistemic stance, and tone, spanning 17 fine-grained types. By pairing these EoBs with world knowledge facts, we generate controlled EoB–query pairs that isolate the effect of linguistic variation. Using this benchmark, we evaluate 16 LLMs that differ in architecture (Llama3, Qwen3, Gemma3), scale (1B-30B parameters), and training stages (base vs instruct). We identify meaningful variations in response behavior across these axes, e.g., that bigger models and instruction models tend to be less context--following than smaller models and base models. We further identify specific EoBs that statistically significantly persuade LMs more consistently than others. Our work reveals systematic patterns in how linguistic framing affects LLM context integration, with implications for prompt engineering and model robustness.

It’s Not What You Say, It’s How You Say It: Evaluating LLM Responses to Expressions of Belief

While guided decoding, especially value-guided methods, has emerged as a cost-effective alternative for controlling language model outputs without re-training models, its effectiveness is limited by the accuracy of the value function. We identify that this inaccuracy stems from a core distributional gap: existing methods train static value functions on trajectories sampled exclusively from the base policy, which inherently confines their training to a narrow and suboptimal view of the potential output space. We propose Iterative Value Refinement, a novel framework designed to bridge this gap. It employs Value Exploration to provide a more comprehensive and robust training signal, complemented by Iterative Self-Refinement, which uses the improved value function from one iteration to guide the generation of higher-quality data for the next. Extensive experiments on text summarization, multi-turn dialogue, and instruction following demonstrate the effectiveness of our framework in aligning language models. Our approach not only achieves alignment but also significantly reduces computational costs by leveraging principled value function optimization for efficient and effective control.

Downloads

Next from ACL 2026

ImReasoner: Improving Memory-based Language Models for Reasoning-in-a-Haystack Tasks

Similar lecture

Domain Pre-training Impact on Representations