CS Paper Radar

Observational Equivalence of LLM and Human Annotation

Kentaro Nakamura, Jing Ling Tan, George Yean

cs.CL ·submitted 2026-08-24 07:23 UTC · v2

Abstract

In this paper, we show that humans and large language models (LLMs) are observationally equivalent in coding texts. Recent LLMs agree with attentive humans at rates comparable to those observed among humans, and cross-LLM disagreement concentrates on texts that also divide humans. We demonstrate these patterns through replications of text-classification tasks from 14 peer-reviewed political science studies, with 13 LLMs, three human coders (Ph.D. students), and 165 crowd coders, independently applying identical codebooks. We further find that clarifying an underspecified codebook increases agreement among humans and between humans and LLMs. Together, these findings point to ambiguity in texts and coding rules as a shared source of disagreement and challenge the presumption that human labels should take precedence whenever humans and LLMs differ. We therefore argue that the central challenge of text annotation is not simply choosing between humans and LLMs, but developing coding rules that minimize ambiguity and accounting for the ambiguity that remains. To this end, we propose using cross-LLM disagreement to identify difficult cases, guide human review, and refine codebooks, and we develop ambiguity-aware bounds for downstream inference without requiring ground truth for every text.

Closest papers

Nearest neighbours by meaning across the whole index.

JEV versus LLMs: Accuracy, Cost and Calibration on Seven Political Science Replications

cs.CL cs.CY ·2026-10-05 ·Steven Denney, Matthew DiGiuseppe
similarity 0.69
Abstract · 71 pages, 2 figures, 14 tables (including appendices). v2: corrected author order in metadata

Large language models (LLMs) annotate and scale political text or constructs by generating text tokens. A new class of models, which TypeSafe markets as "System One" models, instead returns decisions and probability distributions across a user-supplied fixed answer set. A commercial model, JEV, is advertised as having a dramatic cost and speed advantage over traditional LLMs along with better calibrated decisions. As such, it might be useful for social scientists looking to quickly and cost-effectively annotate or scale large corpora of text and have a reliable indicator of a classifier's uncertainty. Yet, the accuracy of these claims and the broader model accuracy in social science text-based tasks are not yet established. In this paper, we do just that and hope to establish the suitability of JEV for social science tasks. We compare JEV with LLMs and human coders from published research, and with a current mid-tier commercial LLM (GPT-6 Luna) and an open-weight alternative (Qwen3.8-27B). We find that JEV matches, or comes close to, the capabilities of both LLMs in a variety of tasks. However, we find no cost advantage over GPT-6 Luna at OpenAI's batch prices. Further, we find that, when each question is asked once, JEV's probabilities are better calibrated than GPT-6 Luna's token probabilities, but not consistently better than Qwen3.8-27B's. We conclude that unless researchers have a need for speed, JEV's only obvious advantage is ease of parsing the underlying choice probabilities.

0.34

CoDA: A Human-LM Collaborative Codebook Development Assistant for Domain-Specific Text Annotation

cs.CL cs.AI ·2026-10-16 ·Jaspreet Ranjit, Swaminathan Chellappa, Augusto Rivas Costante, Chandan Manjunath, Laura Onasch-Vera +3 more
top venuecode/data similarity 0.67
Abstract · In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: System Demonstrations

We present CoDA, an open-source human-LM collaborative codebook development assistant that helps domain experts develop annotation guidelines for annotating new variables in unstructured text with language models. Existing language model annotation systems primarily focus on downstream classification performance, while assuming the availability of high-quality annotation guidelines, or codebooks. In practice, developing these codebooks is time-consuming and challenging, particularly for concepts that require expert knowledge. CoDA addresses this gap by collaborating with domain experts to iteratively develop codebooks through targeted review of incorrect language model predictions and incorporating expert feedback into the evolving codebook. We demonstrate CoDA on two real-world case studies: extracting health indicators from homelessness case management notes and classifying sentiment of wildlife trafficking posts on social media. Across both domains, CoDA produces codebooks that achieve downstream classification performance comparable to expert codebooks while reducing codebook development time from several weeks to two hours and requiring 5X fewer manually annotated examples. Applying CoDA to generate annotations at scale, we identified 2,410 case notes that contained health discussions or referrals and 10,196 social media posts expressing negative sentiment that may warrant investigation for potential wildlife trafficking. By accelerating codebook development, CoDA helps practitioners use language models to extract domain-specific insights from unstructured text at scale.

0.45

Experts Rise Where LLMs Disagree: Using Cross-Model Disagreement to Target Expert Effort in LLM Codebook Revision for Large-Scale Annotation

cs.CL cs.AIcs.HCcs.LG ·2026-09-22 ·Zeyu He, Zhuqian Zhou, Kirk Vanacore, Rene F. Kizilcec, Ting-Hao 'Kenneth' Huang
similarity 0.67
Abstract

Large-scale text annotation brings expert insight to millions of documents, often through a codebook that AI annotators follow. Developing a robust codebook, however, takes months. Large language models (LLMs) could speed this process by applying an early codebook to the data, surfacing cases with strong LLM disagreement, and eliciting expert feedback to address them. We examined three ways experts can provide feedback for LLM codebook revision: (i) editing LLM-generated revisions driven by cross-LLM disagreement (Codebook Verifying), (ii) answering questions about LLM disagreements (Question Answering), and (iii) labeling disagreement cases with rationales (Rationale Labeling). Experiments on thousands of tutoring-session transcripts show that Rationale Labeling yielded the highest LLM-labeling accuracy (64.9%) against expert labels, outperforming the expert-revised codebook (57.8%). The best Question Answering setting also outperformed it (60.5%). Our work shows that LLMs can be used to strategically target expert attention, shortening months of codebook revision to days without sacrificing labeling performance.

0.43

Decomposing and Predicting Annotation Disagreement in Harmful Content Across Languages

cs.CL cs.CYcs.MM ·2026-10-13 ·Irem Ecem Yelkanat, Faeze Ghorbanpour, Alexander Fraser
similarity 0.65
Abstract

Annotator disagreement is increasingly recognized as a meaningful property of subjective tasks, rather than merely noise to be removed. Predicting which items are likely to be contested could help allocate annotation effort and route difficult cases to human review. We propose predicting disagreement from interpretable semantic properties of the text. An LLM answers a set of binary questions covering harmful-content characteristics, disagreement-related cues such as borderline severity and missing context, and language-sensitive cues such as local slurs. A classifier predicts disagreement from these answers, and we analyze which questions contribute most to the predictions. Across four datasets in English, German, and Turkish and three open-weight LLMs, our structured question features outperform direct prompting by 10.5 ROC-AUC points and the strongest text-based baselines by 5.0 points. A compact subset of eight questions performs within 1.5 ROC-AUC points of the full twenty-question set. In leave-one-dataset-out evaluation, models retain roughly 90-95% of their in-domain ROC-AUC. Across models and datasets, questions about derogation, missing context, and decision boundaries serve as the strongest predictors, suggesting that what makes harmful content contested is partly observable in the text.

0.27

Measuring Definition Sensitivity in LLM Annotation of Social Science Concepts

cs.CL cs.CY ·2026-11-09 ·Bogdan Mamaev, Qizhou Wang, Christopher Leckie, Matteo Vergani
published similarity 0.65
Abstract · Accepted at the NeurIPS 2026 Workshop TAE: Can We Trust AI Evaluation? Robustness, Causality, and Risk in Modern AI Assessment

In social science research, text annotation is a fundamental task that is traditionally a costly, hard-to-scale manual process. Researchers have used large language models (LLMs) to perform annotation by writing a definition of a concept into a prompt - a rule that determines whether a label should be attributed to a text depending on its content. However, concepts such as hate speech, fake news, and terrorism typically have several theoretically defensible competing definitions. It is also unknown whether the model actually applied the definition to provide the label or the label was based on the model's pretrained notion of the concept. To test this, we have the model annotate the same texts twice, each time with a different definition. We call the disagreement between the two sets of annotations definition sensitivity. As a control, we paraphrase one definition but leave its requirements unchanged - the disagreement between the annotations with the two wordings becomes the paraphrase floor. Across six concepts and three LLMs, a swap of competing definitions always changes more annotations, on average, than a paraphrase. This shows that the annotations depend on the prompted definition, so any set of annotations should be interpreted together with the definition that produced it. The LLM annotator's definition sensitivity and its paraphrase floor should both be reported in such work.

0.21

Who is the Pronoun Pro? Comparing LLMs and Humans at Number-Disambiguation of 'they'

cs.CL cs.AIcs.HC ·2026-11-03 ·Carlos Hartmann
benchmark similarity 0.65
Abstract · Submitted to journal for publication, will be updated upon publication; figures are greyscale-compatible; raw data, fully reproducible code,…

This study evaluates whether large language models (LLMs) can reliably replace or complement human research assistants in linguistically complex annotation. In a controlled annotation task, 73 human annotators (local students and crowdworkers with and without a background in linguistics) and more than 40 LLM configurations number-disambiguated 174 authentic uses of 'they' from Reddit, including innovative and context-sensitive cases. The best LLM, Kimi K3, reached 94.5% accuracy, above the human majority vote (86.2%) and the mean of the top half of human annotators (78.5%); only the single best human annotator scored higher (95.7%). LLMs also offer substantial practical advantages in cost and speed. However, qualitative error analysis reveals a crucial complementarity: human reasoning in low-agreement cases exposes inconsistencies in the annotation typology and thereby improves the methodological framework itself. The study contributes (i) a benchmark-style human-LLM comparison for a semantically demanding task, (ii) transparent cost and time evidence, and (iii) a methodological argument against a purely replacement-oriented view of automation. LLMs are viable for large-scale annotation support in linguistics, but their responsible use requires human critical oversight. The dataset with its gold-standard labels, all human and LLM annotations, and the fully reproducible code are openly available as a Quarto-based, public repository.

0.27

Evaluating Cultural Reasoning and Robustness in LLMs for Spanish Political Discourse

cs.CL cs.CYcs.SI ·2026-11-16 ·Jorge Valverde-Rebaza, Nathan Garcia-Larios, Karime Ruvalcaba-Perez, Alejandro Rendon, Santiago Sanchez-Ballesteros +1 more
published similarity 0.64
Abstract · Accepted for publication in Latin American Conference on Informatics (CLEI 2026)

Large language models (LLMs) are increasingly used to interpret, generate, and mediate public-facing language, which makes their reliability especially consequential in politically sensitive settings. Although sentiment analysis is a common task in this area, affective meaning in social media discourse is often conveyed through culturally situated and pragmatically indirect cues that are not well captured by label-centric evaluation alone. In this paper, we evaluate the robustness of LLMs for sentiment analysis of Spanish political discourse from social media, with emphasis on culturally situated affective interpretation. We introduce a diagnostic protocol that combines three independent human annotation views into Gold and Ambiguous subsets, and extends the corpus with a Synthetic subset built from sentiment-flipped counterfactuals selected through model-conditioned perplexity and LLM-as-a-judge criteria. Using this framework, we compare proprietary and open-weight models under zero-shot and few-shot prompting, focusing on predictive robustness across human agreement conditions and counterfactual shift. Our results show that annotation ambiguity and counterfactual perturbation reveal failure modes that conventional evaluation on the original corpus tends to obscure. GPT-4.1-mini remains the most stable among the evaluated models, whereas the two smaller open-weight models degrade more markedly under synthetic perturbations, and few-shot prompting provides only limited gains. Overall, the paper contributes a reproducible protocol for robustness-oriented evaluation in political sentiment analysis, an empirical characterization of model behavior under ambiguity and sentiment-flipped counterfactuals, and a publicly available Spanish social media dataset with three-way human sentiment annotations.

0.17

BAIT: Can LLMs Find What Biases a Reader? A Benchmark for Cognitive Bias Triggers in News

cs.CL cs.AI ·2026-10-20 ·Yousra Fettach, Tijl De Bie
benchmark similarity 0.64
Abstract

Persuasive writing exploits cognitive biases, recruiting a reader's mental shortcuts to shape belief, often without the reader noticing. We call the linguistic features exploiting such mental shortcuts cognitive bias triggers. As large language models (LLMs) increasingly mediate how people encounter news, summarizing it, ranking it, and answering questions over it, a natural question is whether they can recognize the features of a text that bias a human reader. We introduce BAIT (Bias Annotation In Text), a resource for detecting cognitive bias triggers in news. BAIT pairs a schema of nine triggers, each specified in an annotator codebook, with 50 English news articles annotated by trained annotators at the article and span levels and adjudicated into a gold standard. On this data we define two tasks, detecting which triggers a text contains and locating the spans that carry them. We then benchmark six frontier LLMs zero-shot and with three or ten in-context examples. The models approach human agreement on which triggers an article contains, but rarely recover its full set of triggers, and they reach less than half of human agreement in locating the passages that carry them. Moreover, they over-predict triggers that come with easy surface cues, and most of the passages they mark for these triggers are not triggers, yet carry the trigger's surface cue far more often than random text. These findings suggest that the evaluated LLMs respond to surface cues rather than to the mechanisms that make a passage manipulative. We release the schema, the codebook, standoff annotations, and the evaluation pipeline.

0.18