United States

Gold&#39;&#39; and &#39;&#39;ground truth&#39;&#39; human-mediated labels have error. This error can escape commonly reported metrics of label quality or obscure questions of accuracy, bias, fairness, and usefulness during model evaluation. This study demonstrates methods for answering such questions even in the context of very low reliabilities from expert humans. We analyze human labels, GPT model ratings, and transformer encoder model ratings of the quality of classroom teaching from two LLM architecture families--encoders and GPT decoders. First, we demonstrate that using standard metrics in the presence of poor labels can mask both label and model quality. The encoder family of models achieve state-of-the-art, even &#39;&#39;super-human&#39;&#39;, results across all classroom annotation tasks using standard metrics. However, evaluation techniques accounting for unreliable labels reveal important flaws, including spurious correlations and nonrandom racial biases across models and humans. We estimate that if models were used in a human-in-the-loop context, the variance contributed by GPT model labels would worsen ratings. These techniques also highlight tasks where encoders could offer 80% reduction in human costs while also reducing bias.

NAACL 2025

All that Glitters&#39;&#39;: Techniques for Evaluations with Unreliable Model and Human Annotations

human-subject application-grounded evaluations

data ethics

model bias/unfairness mitigation

statistical testing for evaluation

model bias/fairness evaluation

ethical considerations in nlp applications

calibration/uncertainty

evaluation methodologies

human-centered evaluation

educational applications

transparency

human-in-the-loop

multi-task learning

metrics

### Welcome to ACL 2023, the 61st Annual Meeting of the Association for Computational Linguistics! 
 The conference will be held in Toronto, Canada, July 9-14, 2023. 
Following the succession of the recent conferences in our field, ACL 2023 will adopt a hybrid format.
While the impact of Covid has considerably diminished in terms of traveling, obtaining visas to Canada
entails a very long process. Moreover, the global economic conditions pose challenges for many individuals to travel to conferences. Recognizing these circumstances, we know many participants may not be
able to attend the conference in person. Therefore, we are committed to providing a great virtual platform
so everyone has the opportunity to interact with other participants and enjoy the conference. Based on the
current registered participants, approxiately 30% have chosen to attend the conference virtually. Whether
you join us in person or virtually, we sincerely hope everyone has a remarkable conference experience. 
This General Chair’s message is where I express my gratitude to the many individuals who have made
enormous contributions to the conference over the past year.

Read [**ACL 2023 General Chair's message**](https://docs.google.com/document/d/1WobYM7norbG4dI48s75HfJoD89qgX5a_F-6U8AteLSA/edit?usp=sharing/) in full.

##### **[Conference Handbook](https://2023.aclweb.org/downloads/acl2023-handbook.pdf)**

ACL 2023

The Association for Computational Linguistics (ACL) is the premier international scientific and professional society for people working on computational problems involving human language, a field often referred to as either computational linguistics or natural language processing.

Gold'' and ''ground truth'' human-mediated labels have error. This error can escape commonly reported metrics of label quality or obscure questions of accuracy, bias, fairness, and usefulness during model evaluation. This study demonstrates methods for answering such questions even in the context of very low reliabilities from expert humans. We analyze human labels, GPT model ratings, and transformer encoder model ratings of the quality of classroom teaching from two LLM architecture families--encoders and GPT decoders. First, we demonstrate that using standard metrics in the presence of poor labels can mask both label and model quality. The encoder family of models achieve state-of-the-art, even ''super-human'', results across all classroom annotation tasks using standard metrics. However, evaluation techniques accounting for unreliable labels reveal important flaws, including spurious correlations and nonrandom racial biases across models and humans. We estimate that if models were used in a human-in-the-loop context, the variance contributed by GPT model labels would worsen ratings. These techniques also highlight tasks where encoders could offer 80% reduction in human costs while also reducing bias.

All that Glitters'': Techniques for Evaluations with Unreliable Model and Human Annotations

poster

### Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Welcome to the 2025 meeting of the Nations of the Americas Chapter of the Association for Computational Linguistics! I am proud to help organize the first NAACL conference to carry the new name of our organization, one that emphasizes inclusion for all of the Americas. I am also pleased to welcome you to Albuquerque, New Mexico, a state whose unique blend of cultural influences will make for an excellent backdrop for NAACL 2025, especially with this year’s special theme on NLP in a Multicultural World. 
**[Continue reading...](https://drive.google.com/file/d/1jX-qGhqVSZZCIrAnJaz798pflu5Irrdn/view?usp=sharing)**

*- Colin Cherry, Google, NAACL 2025 General Chair* 

[![](https://assets.underline.io/markdown_image/1/image/b087f8a4dc5816d6e1a6514e59c59ac3.png)](https://drive.google.com/file/d/1T96GzPqObXrMln2BMByCSXTSizjTg69P/view?usp=sharing)

You need to log in with the email address you registered with. Access credentials have been sent to your email. 

Please be sure to check your spam and other email folders if you do not see an email confirmation right away.

Please log in to explore this event.

To access NAACL 2025 event page you are required to register. Please follow [**this link**](https://2025.naacl.org/registration/registration/) to register. Access will depend on your registration type.

Please register!

Welcome to 2025 Annual Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics

Recent advancements in Large Language Models (LLMs) have significantly enhanced interactions between users and models. These advancements concurrently underscore the need for rigorous safety evaluations due to the manifestation of social biases, which can lead to harmful societal impacts. Despite these concerns, existing benchmarks may overlook the intrinsic weaknesses of LLMs, which can generate biased responses even with simple adversarial instructions. To address this critical gap, we introduce a new benchmark, Fairness Benchmark in LLM under Extreme Scenarios (FLEX), designed to test whether LLMs can sustain fairness even when exposed to prompts constructed to induce bias. To thoroughly evaluate the robustness of LLMs, we integrate prompts that amplify potential biases into the fairness assessment. Comparative experiments between FLEX and existing benchmarks demonstrate that traditional evaluations may underestimate the inherent risks in models. This highlights the need for more stringent LLM evaluation benchmarks to guarantee safety and fairness.

FLEX: A Benchmark for Evaluating Robustness of Fairness in Large Language Models

Existing studies have shown that AI-generated images tend to reinforce social biases, including those related to race and gender. However, no studies have investigated weight bias, or fatphobia, in AI-generated images. This study utilizes DALL-E~3 to determine the extent to which anti-fat and pro-thin biases are present in AI-generated images, and examines stereotypical associations between moral character and body weight. Four-thousand images are generated using twenty pairs of positive and negative textual prompts. These images are then manually labeled with weight information and analyzed to determine the extent to which they reflect fatphobia. The findings and their impact are discussed and related to existing research on weight bias.

Decoding Fatphobia: Examining Anti-Fat and Pro-Thin Bias in AI-Generated Images

The proliferation of large language models has raised growing concerns about their misuse, particularly in cases where AI-generated text is falsely attributed to human authors. Machine-generated content detectors claim to effectively identify such text under various conditions and from any language model. This paper critically evaluates these claims by assessing several popular detectors (RADAR, Wild, T5Sentinel, Fast-DetectGPT, PHD, LogRank, Binoculars) on a range of domains, datasets, and models that these detectors have not previously encountered. We employ various prompting strategies to simulate practical adversarial attacks, demonstrating that even moderate efforts can significantly evade detection. We emphasize the importance of the true positive rate at a specific false positive rate (TPR@FPR) metric and demonstrate that these detectors perform poorly in certain settings, with TPR@.01 as low as 0%. Our findings suggest that both trained and zero-shot detectors struggle to maintain high sensitivity while achieving a reasonable true positive rate.

A Practical Examination of AI-Generated Text Detectors for Large Language Models

There has been significant prior work using templates to study bias against demographic attributes in MLMs. However, these have limitations: they overlook random variability of templates and target concepts analyzed, assume equality amongst templates, and overlook bias quantification. Addressing these, we propose a systematic statistical approach to assess bias in MLMs, using mixed models to account for random effects, pseudo-perplexity weights for sentences derived from templates and quantify bias using statistical effect sizes. Replicating prior studies, we match on bias scores in magnitude and direction with small to medium effect sizes. Next, we explore the novel problem of gender bias in the context of *personality* and *character* traits, across seven MLMs (base and large). We find that MLMs vary; ALBERT is unbiased for binary gender but the most biased for non-binary *neo*, while RoBERTa-large is the most biased for binary gender but shows small to no bias for *neo*. There is some alignment of MLM bias and findings in psychology (human perspective) - in *agreeableness* with RoBERTa-large and *emotional stability* with BERT-large. There is general agreement for the remaining 3 personality dimensions: both sides observe at most small differences across gender. For character traits, human studies on gender bias are limited thus comparisons are not feasible.

Robust Bias Detection in MLMs and its Application to Human Trait Ratings

With the increasing adoption of large language models (LLMs) in education, concerns about inherent biases in these models have gained prominence. We evaluate LLMs for bias in the personalized educational setting, specifically focusing on the models’ roles as “teachers.” We reveal significant biases in how models generate and select educational content tailored to different demographic groups, including race, ethnicity, sex, gender, disability status, income, and national origin. We introduce and apply two bias score metrics—Mean Absolute Bias (MAB) and Maximum Difference Bias (MDB)—to analyze 9 open and closed state-of-the-art LLMs. Our experiments, which utilize over 17,000 educational explanations across multiple difficulty levels and topics, uncover that models potentially harm student learning by both perpetuating harmful stereotypes and reversing them. We find that bias is similar for all frontier models, with the highest MAB along income levels while MDB is highest relative to both income and disability status. For both metrics, we find the lowest bias exists for sex/gender and race/ethnicity.

LLMs are Biased Teachers: Evaluating LLM Bias in Personalized Education

Preference alignment via reward models helps build safe, helpful, and reliable large language models (LLMs). However, subjectivity in preference judgments and the lack of representative sampling in preference data collection can introduce new biases, hindering reward models' fairness and equity. In this work, we introduce a framework for evaluating dialect biases in reward models and conduct a case study on biases against African American Language (AAL) through several experiments comparing reward model preferences and behavior on paired White Mainstream English (WME) and both machine-translated and human-written AAL corpora. We show that reward models are less aligned with human preferences when processing AAL texts vs. WME ones (-4% accuracy on average), frequently disprefer AAL-aligned texts vs. WME-aligned ones, and steer conversations toward WME, even when prompted with AAL texts. Our findings provide a targeted analysis of anti-AAL biases at a relatively understudied stage in LLM development, highlighting representational harms and ethical questions about the desired behavior of LLMs concerning AAL.

Rejected Dialects: Biases Against African American Language in Reward Models

Reinforcement learning with human feedback (RLHF) is shown to largely benefit from precise reward models (RMs). However, recent studies in reward modeling schemes are skewed towards English, limiting the applicability of RLHF in multilingual alignments. In this work, we investigate the cross-lingual transfer of RMs trained in diverse languages, primarily from English. Our experimental results demonstrate the strong cross-lingual transfer of English RMs, exceeding target language RMs by 3~4% average increase in Multilingual RewardBench. Furthermore, we analyze the cross-lingual transfer of RMs through the representation shifts. Finally, we perform multilingual alignment to exemplify how cross-lingual transfer in RM propagates to enhanced multilingual instruction-following capability.

Cross-lingual Transfer of Reward Models in Multilingual Alignment

Reranking, or scoring a list of prediction candidates from a machine translation system with an external scoring model and returning the highest-scoring candidate, remains a simple and effective method for improving prediction quality. However, reranking with high quality scoring models can add substantial computational cost to the translation pipeline, which we address in this work by framing list reranking as a Bayesian optimization (BayesOpt) problem over the candidate list, where unknown scores are modeled with a Gaussian process. This algorithm scores candidates iteratively, choosing next candidates by balancing between exploration, choosing to score those that differ from candidates already scored, and exploitation, choosing to score those that resemble high-scoring candidates. This procedure finds high-scoring candidates while scoring only a fraction of the candidates list; given candidate lists of 200 random samples (before deduplication), our method achieves the same CometKiwi score using only 70 scoring evaluations on average compared to scoring a random subset of 180 candidates. We also propose multi-fidelity BayesOpt for list reranking, where scores obtained from a noisier but cheaper proxy scoring model are incorporated into the search process. We show that well-trained distilled proxy scorers can further improve the performance of BayesOpt.

A Bayesian Optimization Approach to Machine Translation Reranking

Large language models (LLMs) have become standard for natural language generation tasks, with instruction-tuning enhancing their capabilities. However, the lack of instruction-tuning datasets in languages other than English limits their application to diverse languages. To address this, researchers have adapted English-centric LLMs to other languages by appending English tuning data with its translated pair, from which we observe negative interference between the two. To resolve this, our contribution is identifying English as an internal pivot language, based on which we disentangle the roles of English and target language data in training. Specifically, we first design two roles as pivoted objectives, and also propose to regulate between the two, to better generalize for under-represented languages. Experiments across various languages demonstrate the effectiveness of our approach on multiple benchmarks. The code is publicly available for further exploration.

PROM: Pivoted and Regulated Optimization for Multilingual Instruction Learning

Existing approaches to multilingual text detoxification are hampered by the scarcity of parallel multilingual datasets. In this work, we introduce a pipeline for the generation of multilingual parallel detoxification data. We also introduce SynthDetoxM, a manually collected and synthetically generated multilingual parallel text detoxification dataset comprising 16,000 high-quality detoxification sentence pairs across German, French, Spanish and Russian. The data was sourced from different toxicity evaluation datasets and then rewritten with nine modern open-source LLMs in few-shot setting. Our experiments demonstrate that models trained on the produced synthetic datasets have superior performance to those trained on the human-annotated MultiParaDetox dataset even in data limited setting. Models trained on SynthDetoxM~outperform all evaluated LLMs in few-shot setting. We release our dataset and code to help further research in multilingual text detoxification.

Downloads

Next from NAACL 2025

FLEX: A Benchmark for Evaluating Robustness of Fairness in Large Language Models

Similar lecture

Can Large Language Models Be an Alternative to Human Evaluations?