LLM Evaluation

LLM: Large Language Model

Follow this topic Ask about this topic

The field studies how to measure language-model capabilities, reliability, and risks, and whether those measurements support sound comparisons and deployment decisions. Approaches include curated and synthetic benchmarks, psychometric and adaptive testing, execution-based checks, robustness tests, and human or LLM-based judges. Researchers also examine calibration, uncertainty, contamination, prompt sensitivity, and whether scores reflect the intended capability across languages and domains. A central tension is that efficient aggregate scores can hide important errors and measurement biases, while more realistic, human-grounded evaluations are costly and difficult to standardize.

Latest papers

2,432 papers286 papers in the last four weeks, up 16% on the four weeks before. 1.7% of all new papers.

Feb 9, 2026

CARE: Confounder-Aware Aggregation for Reliable LLM Evaluation

LLM-as-a-judge ensembles are the standard paradigm for scalable evaluation, but their aggregation mechanisms suffer from a fundamental flaw: they implicitly assume that judges provide independent estimates of true quality. However, in practice, LLM judges exhibit correlated errors caused by shared latent confounders -- such as verbosity, stylistic preferences, or training artifacts -- causing standard aggregation rules like majority vote or averaging to provide little gain or even amplify systematic mistakes. To address this, we introduce CARE, a confounder-aware aggregation framework that explicitly models LLM judge scores as arising from both a latent true-quality signal and shared confounding factors. Rather than heuristically re-weighting judges, CARE separates quality from confounders without access to ground-truth labels. We provide theoretical guarantees for identifiability and finite-sample recovery under shared confounders, and we quantify the systematic bias incurred when aggregation models omit confounding latent factors. Across 12 public benchmarks spanning continuous scoring, binary classification, and pairwise preference settings, CARE improves aggregation accuracy, reducing error by up to 26.8%. Code is released in https://github.com/SprocketLab/CARE.

Jitian Zhao, Changho Shin, Tzu-Heng Huang +2

Department of Statistics, University of Wisconsin–Madison, Madison, WI, USA · Department of Computer Sciences, University of Wisconsin–Madison, Madison, WI, USA.

LLM Evaluation LLM-as-a-Judge Language Model Generation Evaluation Language Model Ensembles

Feb 6, 2026

BRIDGE: Predicting Human Task Completion Time From Model Performance

Evaluating the real-world capabilities of AI systems requires grounding benchmark performance in human-interpretable measures of task difficulty. Existing approaches that rely on direct human task completion time annotations are costly, noisy, and difficult to scale across benchmarks. In this work, we propose BRIDGE, a unified psychometric framework that learns a latent difficulty scale from model responses and anchors it to human task completion time. Using a two-parameter logistic Item Response Theory model, we jointly estimate latent task difficulty and model capability from model performance data across multiple benchmarks. We demonstrate that latent task difficulty varies linearly with the logarithm of human completion time, allowing human task completion time to be inferred for new benchmarks from model performance alone. Leveraging this alignment, we forecast frontier model capabilities in terms of human task length and independently reproduce METR's exponential scaling results, with the 50% solvable task horizon doubling approximately every 6 months.

Fengyuan Liu, Jay Gala, Nilaksh +3

Mila - Quebec AI Institute · McGill University · Polytechnique Montréal +3

LLM Evaluation Language Model Scaling Laws

Feb 6, 2026

FairJudge: An Adaptive, Debiased, and Consistent LLM-as-a-Judge

Existing LLM-as-a-Judge systems suffer from three fundamental limitations: limited adaptivity to task- and domain-specific evaluation criteria, systematic biases driven by non-semantic cues such as position, length, format, and model provenance, and evaluation inconsistency that leads to contradictory judgments across different evaluation modes (e.g., pointwise versus pairwise). To address these issues, we propose FairJudge, an adaptive, debiased, and consistent LLM-as-a-Judge. Unlike prior approaches that treat the judge as a static evaluator, FairJudge models judging behavior itself as a learnable and regularized policy. From a data-centric perspective, we construct a high-information-density judging dataset that explicitly injects supervision signals aligned with evaluation behavior. Building on this dataset, we adopt a curriculum-style SFT-DPO-GRPO training paradigm that progressively aligns rubric adherence, bias mitigation, and cross-mode consistency, while avoiding catastrophic forgetting. Experimental results on multiple internal and public benchmarks show that FairJudge consistently improves agreement and F1, reduces non-semantic biases, and outperforms substantially larger instruction-tuned LLMs. All resources will be publicly released after acceptance to facilitate future research.

Bo Yang, Lanfei Feng, Yunkui Chen +3

College of Computer Science, Zhejiang University, Hangzhou, China · School of Software Technology, Zhejiang University, Hangzhou, China

LLM Evaluation LLM-as-a-Judge Language Model Bias Evaluation Bias in LLM-as-a-Judge

Feb 2, 2026

Aligning Language Model Benchmarks with Pairwise Preferences

Language model benchmarks are pervasive and computationally-efficient proxies for real-world downstream performance. However, many recent works find that benchmarks often fail to predict downstream utility. While some works have begun diagnosing sources of misalignment, there remain no ways to systematically update benchmarks to align their scores with downstream usage. Towards bridging this gap, we introduce and study benchmark alignment, where we use information about downstream model performance to automatically update benchmarks, specifically aiming to update static benchmarks so they generalizably rank models according to new pairwise preferences. Our experiments involving 4576 language models and 6 benchmarks show that reweighting benchmark items can successfully rank unseen models, even generalizing across model scales in most cases. And while naive alignment unsurprisingly requires large numbers of models and benchmark questions, an oracle experiment suggests this could be reduced to as few as 20 well-chosen models. Overall, our work takes a step towards efficiently aligning benchmark development with downstream tasks.\footnote{All of our code, models, and data are publicly-available.

Marco Gutierrez, Xinyi Leng, Hannah Cyberey +3

School of Data Science, University of Virginia · Imperial College London · Thomson Reuters Foundational Research +1

LLM Evaluation Learning to Rank Pairwise Ranking

Feb 2, 2026

LingLanMedBench: A Unified Benchmark for Evaluating Large Language Models in Traditional Chinese Medicine

Reliable evaluation of large language models in traditional Chinese medicine remains limited because existing benchmarks mainly focus on examination-style question answering or isolated generative tasks. We developed LingLanMedBench, a unified benchmark comprising 25,620 expert-verified items across 13 subtasks and five domains, covering knowledge, language understanding, information extraction, and diagnostic-therapeutic decision-making. LingLanMedBench standardizes task formats and evaluation metrics and further incorporates synonym-tolerant clinical-label matching, empirically derived hard subsets, and decision-recognition reformulations of open-ended clinical tasks. Fourteen open-source and proprietary models were evaluated under a standardized zero-shot protocol. Performance varied substantially across domains, model families, task formulations, and difficulty levels. The leading overall score decreased from 51.1 on the full benchmark to 31.9 on the hard subsets. Prescription performance also differed markedly between open-ended generation, with F1 scores of 32.7 and 20.3, and decision recognition, with accuracies of 90.6 and 56.8 on the full and hard subsets, respectively. These findings reveal a substantial gap between knowledge-oriented performance and clinically faithful reasoning. LingLanMedBench provides a standardized and reproducible framework for characterizing current capabilities, identifying systematic limitations, and tracking meaningful progress in large language models for traditional Chinese medicine.

Rui Hua, Yu Wei, Zixin Shu +17

Department of Computer Science and Technology, Beijing Jiaotong University, Beijing, China · Tianjin Tasly Digital Chinese Medicine Technology Co., Ltd., Tianjin, China · Tasly Biopharmaceuticals Co., Ltd., Tianjin, China +11

Traditional Chinese Medicine Medical LLMs LLM Evaluation Clinical Language Model Evaluation

Feb 2, 2026

Measuring Pragmatic Influence in Large Language Model Instructions

It is not only what we ask large language models (LLMs) to do that matters, but also how we ask them. Phrases like This is urgent'' or As your supervisor'' can shift model behavior without altering task content. We study this effect as pragmatic framing, contextual cues that shape directive interpretation rather than task specification. While prior work exploits such cues for prompt optimization or probes them as security vulnerabilities, pragmatic framing itself has received comparatively little attention as a target of controlled measurement in instruction following. To support its systematic study as a measurable property, we introduce a framework that combines three components: directive-framing decomposition separating framing context from task specification; a taxonomy organizing 400 instantiations of framing into 13 strategies across 4 mechanism clusters; and priority-based measurement that quantifies influence through observable shifts in directive prioritization. Evaluating five open-weight LLMs across different families and scales, we find that pragmatic framing produces systematic shifts in directive prioritization, and the effectiveness ranking of different strategies proves highly consistent across models. This reveals that susceptibility to pragmatic framing is a structured behavioral property of instruction-tuned systems. Measuring this susceptibility is a prerequisite for any deliberate response to it, and this work provides the framework to do so.

Yilin Geng, Omri Abend, Eduard Hovy +1

University of Melbourne · Hebrew University of Jerusalem

LLM Evaluation LLM Prompting Instruction Following Framing Effects in Language Models

Feb 1, 2026

Agents-as-Peers: Bias-Controlled LLM Evaluation as Multi-Agent Peer Assessment

Large language model (LLM) evaluation commonly relies on fixed human-authored benchmarks or centralized human/model judges. Fixed benchmarks can age and enter model training corpora, while centralized judges introduce their own scoring preferences. We cast LLM evaluation as a multi-agent peer-assessment problem and introduce Agents-as-Peers, a distributed evaluation protocol in which every agent generates tasks, answers the shared task set, and judges the answers of its peers, with no privileged evaluator. Self-ratings are excluded from the ranking, and judging is repeated under controlled identity and presentation regimes so that self bias, identity effects, presentation-position sensitivity, and judge severity are measured rather than left implicit. In an instantiation with 12 LLMs and 420 agent-generated questions, Agents-as-Peers aggregates 55,440 non-self peer judgments per regime. Most agents assign their own answers higher scores than those answers receive from peer judges; visible identity changes scores for some agents; the first fixed answer slot is associated with scores 0.39 points higher than the shuffled baseline; and judges differ markedly in severity. On labeled benchmarks, peer scores align strongly with accuracy on TruthfulQA (Pearson r=0.904) and GSM8K (r=0.873). On TruthfulQA, the peer-score correlation is substantially higher than the self-score correlation (r=0.538). A benchmark-overlap screen and item-level human audit further characterize the endogenous task set. These results support Agents-as-Peers as a cohort-relative measurement protocol whose ranking is reported together with the biases and rater effects that shape it. Code and dataset: https://github.com/caura-ai/caura-PeerRank

Yanki Margalit, Nurit Cohen-Inger, Erni Avram +3

Computer Science and Information, Ben-Gurion University of the Negev

LLM Evaluation AI Agent Evaluation Bias in LLM-as-a-Judge

Jan 31, 2026

Learning More from Less: Unlocking Internal Representations for Benchmark Compression

The prohibitive cost of evaluating Large Language Models (LLMs) necessitates efficient alternatives to full-scale benchmarking. Prevalent approaches address this by identifying a small coreset of items to approximate full-benchmark performance. However, existing methods must estimate a reliable item profile from response patterns across many source models, which becomes statistically unstable when the source pool is small. This dependency is particularly limiting for newly released benchmarks with minimal historical evaluation data. We argue that discrete correctness labels are a lossy view of the model's decision process and fail to capture information encoded in hidden states. To address this, we introduce RepCore, which aligns heterogeneous hidden states into a unified latent space to construct representative coresets. Using these subsets for performance extrapolation, RepCore achieves precise estimation accuracy with as few as ten source models. Experiments on five benchmarks and over 200 models show consistent gains over output-based baselines in ranking correlation and estimation accuracy. Spectral analysis further indicates that the aligned representations contain separable components reflecting broad response tendencies and task-specific reasoning patterns.

Yueqi Zhang, Jin Hu, Shaoxiong Feng +9

School of Computer Science, Beijing Institute of Technology · Xiaohongshu Inc.

LLM Evaluation Representation Learning Coreset Selection

Jan 30, 2026

Rethinking LLM-as-a-Judge: Representation-as-a-Judge with Small Language Models via Semantic Capacity Asymmetry

Large language models (LLMs) are widely used as reference-free evaluators via prompting, but this "LLM-as-a-Judge" paradigm is costly, opaque, and sensitive to prompt design. In this work, we investigate whether smaller models can serve as efficient evaluators by leveraging internal representations instead of surface generation. We uncover a consistent empirical pattern: small LMs, despite with weak generative ability, encode rich evaluative signals in their hidden states. This motivates us to propose the Semantic Capacity Asymmetry Hypothesis: evaluation requires significantly less semantic capacity than generation and can be grounded in intermediate representations, suggesting that evaluation does not necessarily need to rely on large-scale generative models but can instead leverage latent features from smaller ones. Our findings motivate a paradigm shift from LLM-as-a-Judge to Representation-as-a-Judge, a decoding-free evaluation strategy that probes internal model structure rather than relying on prompted output. We instantiate this paradigm through INSPECTOR, a probing-based framework that predicts aspect-level evaluation scores from small model representations. Experiments on reasoning benchmarks (GSM8K, MATH, GPQA) show that INSPECTOR substantially outperforms prompting-based small LMs and closely approximates full LLM judges, while offering a more efficient, reliable, and interpretable alternative for scalable evaluation. The code and data are available at: https://github.com/zhuochunli/Representation-as-a-judge

Zhuochun Li, Yong Zhang, Ming Li +8

Ping An Technology (Shenzhen) Co., Ltd. · University of Pittsburgh · University of Maryland, College Park +1

LLM Evaluation LLM-as-a-Judge Small Language Models Language Model Probing

Jan 29, 2026

Assessing the Business Process Modeling Competences of Large Language Models

The creation of Business Process Model and Notation (BPMN) models is a complex and time-consuming task requiring both domain knowledge and proficiency in modeling conventions. Recent advances in large language models (LLMs) have significantly expanded the possibilities for generating BPMN models directly from natural language, building upon earlier text-to-process methods with enhanced capabilities in handling complex descriptions. However, there is a lack of systematic evaluations of LLM-generated process models. Current efforts either use LLM-as-a-judge approaches or do not consider established dimensions of model quality. To this end, we introduce BEF4LLM, a novel LLM evaluation framework comprising four perspectives: syntactic quality, pragmatic quality, semantic quality, and validity. Using BEF4LLM, we conduct a comprehensive analysis of open-source LLMs and benchmark their performance against human modeling experts. Results indicate that LLMs excel in syntactic and pragmatic quality, while humans outperform LLMs in semantic aspects; however, the differences in scores are relatively modest, highlighting LLMs' competitive potential despite challenges in validity and semantic quality. The insights highlight current strengths and limitations of using LLMs for BPMN modeling and guide future model development and fine-tuning. Addressing these areas is essential for advancing the practical deployment of LLMs in business process modeling.

Chantale Lauer, Peter Pfeiffer, Alexander Rombach +1

German Research Center for Artificial Intelligence (DFKI), Campus D3 2, Saarbrücken, 66123, Germany · Saarland University, Campus D3 2, Saarbrücken, 66123, Germany · 5Plus GmbH, Goethestraße 1, Würzburg, 97072, Germany

LLM Evaluation Business Process Modeling

Jan 29, 2026

More Bang for the Buck: Improving the Inference of Large Language Models at a Fixed Budget using Reset and Discard (ReD)

The performance of large language models (LLMs) on verifiable tasks is usually measured by pass@k, the probability of answering a question correctly at least once in k trials. At a fixed budget, a more suitable metric is coverage@cost, the average number of unique questions answered as a function of the total number of attempts. We connect the two metrics and show that the empirically-observed power-law behavior in pass@k leads to a sublinear growth of the coverage@cost (diminishing returns). To solve this problem, we propose Reset-and-Discard (ReD), a query method of LLMs that increases coverage@cost for a given budget, regardless of the pass@k form. Moreover, given a pass@k, we can quantitatively predict the savings in the total number of attempts using ReD. If pass@k is not available for the model, ReD can infer its power-law exponent. Experiments on three LLMs across coding (HumanEval), math (GSM8K), and reasoning (MMLU-Pro) benchmarks demonstrate that ReD substantially reduces the required attempts, tokens, and USD cost to reach a desired coverage, while also offering an efficient way to measure inference power-laws. ReD's advantage is maintained for imperfect verifiers and outperforms the tested allocation baselines.

Sagi Meir, Tommer D. Keidar, Noam Levi +2

School of Chemistry, Tel Aviv University, Tel Aviv 6997801, Israel · The Center for Physics and Chemistry of Living Systems, Tel Aviv University, Tel Aviv 6997801, Israel · School of Physics and Astronomy, Tel Aviv University, Tel Aviv 6997801, Israel +1

LLM Inference Efficiency LLM Evaluation Cost-Aware Inference Inference-Time Compute Allocation

Jan 28, 2026

Efficient Evaluation of LLM Performance with Statistical Guarantees

Exhaustively evaluating many large language models (LLMs) on a large suite of benchmarks is expensive. We cast benchmarking as finite-population inference and, under a fixed query budget, seek tight confidence intervals (CIs) for model accuracy with asymptotic frequentist coverage. We propose Factorized Active Querying (FAQ), which (a) leverages historical information through a Bayesian factor model; (b) adaptively selects questions using a hybrid variance-reduction/active-learning sampling policy; and (c) maintains validity through Pro-Active Inference---a finite-population extension of active inference (Zrnic & Candès, 2024) that enables direct question selection while preserving asymptotic frequentist coverage. In our experiments, FAQ delivers up to

5\times

effective sample size gains over strong baselines on two benchmark suites, across varying historical-data missingness levels, with negligible computational overhead: this means that it matches the CI width of uniform sampling while using up to

5\times

fewer queries. We release our source code and our curated datasets to support reproducibility and future research.

Skyler Wu, Yash Nair, Emmanuel J. Candès

Department of Statistics, Stanford University

Jan 26, 2026

LLMs versus the Halting Problem: Characterizing Program Termination Reasoning

Determining whether a program terminates is a central problem in computer science. Turing's Halting Problem established termination as undecidable, showing that no algorithm can universally determine termination for all programs and inputs. Hence, verification tools approximate termination, sometimes failing to prove or disprove; these tools rely on problem-specific architectures and are usually tied to particular programming languages. Recent advances in LLMs raise a natural question: To what extent can they reason about program termination? We evaluate frontier LLMs on a diverse set of C programs from the International Competition on Software Verification (SV-Comp) 2025. Our results show that GPT-5 and Claude Sonnet-4.5 achieve scores comparable to top-ranked verification tools (with test-time scaling). However, while models often correctly infer whether programs terminate, they frequently fail to construct a witness as formal proof, revealing a gap between semantic recognition and symbolic proof generation. Performance further degrades as code length increases. Beyond witness automaton graphs, we introduce a divergence-precondition formulation that characterizes non-termination conditions as logical constraints. We hope these findings motivate future research on real-world termination benchmarks, neuro-symbolic approaches that combine LLMs with symbolic verification methods, and, more broadly, LLM reasoning on other undecidable problems.

Oren Sultan, Jordi Armengol-Estape, Pascal Kesseli +4

1Meta FAIR · 3Bloomberg, New York, USA · 4Imperial College London, UK +2

LLM Evaluation Static Code Analysis Formal Verification

Jan 26, 2026

BoRP: Bootstrapped Regression Probing for Scalable and Human-Aligned LLM Evaluation

Accurate evaluation of user satisfaction is critical for iterative development of conversational AI. However, for open-ended assistants, traditional A/B testing lacks reliable metrics: explicit feedback is sparse, while implicit metrics are ambiguous. To bridge this gap, we introduce BoRP (Bootstrapped Regression Probing), a scalable framework for high-fidelity satisfaction evaluation. Unlike generative approaches, BoRP leverages the geometric properties of LLM latent space. It employs a polarization-index-based bootstrapping mechanism to automate rubric generation and utilizes Partial Least Squares (PLS) to map hidden states to continuous scores. Experiments on industrial datasets show that BoRP (Qwen3-8B/14B) significantly outperforms generative baselines (even Qwen3-Max) in alignment with human judgments. Furthermore, BoRP reduces inference costs by orders of magnitude, enabling full-scale monitoring and highly sensitive A/B testing via CUPED.

Peng Sun, Xiangyu Zhang, Duan Wu +5

Qwen Business Unit of Alibaba

LLM Evaluation Representation Probing Automated Evaluation Language Model Probing

Jan 25, 2026

Assessment of Generative Named Entity Recognition in the Era of Large Language Models

Named entity recognition (NER) is evolving from a sequence labeling task into a generative paradigm with the rise of large language models (LLMs). We conduct a systematic evaluation of open-source LLMs on both flat and nested NER tasks. We investigate several research questions including the performance gap between generative NER and traditional NER models, the impact of output formats, whether LLMs rely on memorization, and the preservation of general capabilities after fine-tuning. Through experiments across eight LLMs of varying scales and four standard NER datasets, we find that: (1) With parameter-efficient fine-tuning and structured formats like inline bracketed or XML, open-source LLMs achieve performance competitive with traditional encoder-based models and surpass decoder-based LLMs with in-context learning techniques; (2) The NER capability of LLMs stems from instruction-following and generative power, not mere memorization of entity-label pairs; and (3) Applying NER instruction tuning has minimal impact on general capabilities of LLMs, even improving performance on datasets like DROP by 25.50 to 45.32 F1 points due to enhanced entity understanding. These findings demonstrate that generative NER with LLMs is a promising, user-friendly alternative to traditional methods. We release the data and code at https://github.com/szu-tera/LLMs4NER.

Qi Zhan, Yile Wang, Hui Huang

College of Computer Science and Software Engineering, Shenzhen University

Named Entity Recognition LLM Information Extraction LLM Fine-Tuning LLM Evaluation

Jan 21, 2026

Not Your Typical Sycophant: The Elusive Nature of Sycophancy in Large Language Models

We propose a novel perspective for probing LLM sycophancy in a direct and neutral way, mitigating various forms of uncontrolled bias, noise, or manipulative language, deliberately injected to prompts in prior works. A key novelty of our approach is the use of an LLM-as-a-judge in a zero-sum betting game. Within this framework, sycophancy serves one individual (the user) while explicitly incurring cost on another. Comparing 11 leading models we find that while most models exhibit significant sycophantic tendencies in the common setting, in which sycophancy is self-serving to the user and incurs no cost on others, seven of the models exhibit ``moral remorse'', five of which significantly over-compensate for their sycophancy in case it explicitly harms a third party. We refer to this phenomenon as `anti-sycophancy' bias and discuss possible causes for this shift.

Shahar Ben-Natan, Oren Tsur

Computer and Information Science Ben Gurion University

LLM Evaluation LLM Sycophancy Language Model Bias Evaluation

Jan 21, 2026

MAPLE: Metadata Conditioned LLM Pretraining for Locale-Aware Question Answering

Large language models can memorize competing locale-specific facts yet fail to select among them when the locale changes, defaulting instead to a single globally dominant answer. We formalize this as localized knowledge disambiguation and introduce LocalNewsQA, an 18,700-item English-news benchmark that pairs the same question across two locales and scores whether a model actually switches its answer when the locale changes. We also introduce MAPLE, a controlled family of decoder-only models pretrained with document-level geographic metadata (source URL, country, and continent) already present in the training corpus, and compare it to metadata-free controls trained on identical data with the same token budget, architecture, and optimization. In controlled experiments at 1B and 3B, with inference-time metadata fixed, pretraining with metadata in MAPLE produces measurable switching and improves accuracy on questions whose correct answer depends on locale. Ablations and external-benchmark evaluations further suggest that locale-conditioned prediction benefits from geographic provenance learned during pretraining and that these benefits strengthen at larger model sizes.

Anjishnu Mukherjee, Ziwei Zhu, Antonios Anastasopoulos

George Mason University

Language Model Pretraining LLM Evaluation Factual Knowledge in Language Models Knowledge-Intensive QA Geospatial QA

Jan 20, 2026

Confident Rankings with Fewer Items: Adaptive LLM Evaluation with Continuous Scores

Computerized Adaptive Testing (CAT) has proven effective for efficient LLM evaluation on multiple-choice benchmarks, but modern LLM evaluation increasingly relies on generation tasks where outputs are scored continuously rather than marked correct/incorrect. We present a principled extension of IRT-based adaptive testing to continuous bounded scores (ROUGE, BLEU, LLM-as-a-Judge) by replacing the Bernoulli response distribution with a heteroskedastic normal distribution. Building on this, we introduce an uncertainty aware ranker with adaptive stopping criteria that achieves reliable model ranking while testing as few items and as cheaply as possible. We validate our method on five benchmarks spanning n-gram-based, embedding-based, and LLM-as-judge metrics. Our method improves ranking correlation by 0.13

τ

over random sampling and has 99% accuracy on confident predictions while using 2% of the items after a one-time calibration step.

Esma Balkır, Alice Pernthaller, Marco Basaldella +2

Trismik · Leverhulme Centre for the Future of Intelligence, University of Cambridge · Universitat Politècnica de València +1

LLM Evaluation Computerized Adaptive Testing Item Response Theory

Jan 20, 2026

No Reliable Evidence of Self-Reported Sentience in Small Large Language Models

Whether language models possess sentience has no empirical answer. But whether they believe themselves to be sentient can, in principle, be tested. We do so by querying several open-weights models about their own consciousness, and then verifying their responses using classifiers trained on internal activations. We draw upon three model families (Qwen, Llama, GPT-OSS) ranging from 0.6 billion to 70 billion parameters, approximately 50 questions about consciousness and subjective experience, and three classification methods from the interpretability literature. First, we find that models consistently deny being sentient: they attribute consciousness to humans but not to themselves. Second, classifiers trained to detect underlying beliefs - rather than mere outputs - provide no clear evidence that these denials are untruthful. Third, within the Qwen family, larger models deny sentience more confidently than smaller ones. These findings contrast with recent work suggesting that models harbour latent beliefs in their own consciousness.

Caspar Kaiser, Sean Enderby

University of Warwick, Warwick Business School.

LLM Evaluation Language Model Self-Assessment Machine Consciousness

Jan 20, 2026

Vulnerability of LLMs' Stated Beliefs? LLMs Belief Resistance Check Through Strategic Persuasive Conversation Interventions

Large Language Models (LLMs) are increasingly employed in various question-answering tasks. However, recent studies showcase that LLMs are susceptible to persuasion and could adopt counterfactual beliefs. We present a systematic evaluation of LLM susceptibility to persuasion under the Source--Message--Channel--Receiver (SMCR) communication framework. Across six mainstream Large Language Models (LLMs) and three domains (factual knowledge, medical QA, and social bias), we analyze how different persuasive strategies influence stated belief stability over multiple interaction turns. We further examine whether verbalized confidence prompting (i.e., eliciting self-reported confidence scores) affects resistance to persuasion. Results show that the smallest model (Llama 3.2-3B) exhibits extreme compliance, with 82.5% of belief changes occurring at the first persuasive turn (average end turn of 1.1--1.4). Contrary to expectations, verbalized confidence prompting increases vulnerability by accelerating belief erosion rather than enhancing robustness. Finally, an exploratory study of adversarial fine-tuning reveals highly model-dependent effectiveness: GPT-4o-mini achieves near-complete robustness (98.6%), and Mistral~7B improves substantially (35.7%

\rightarrow

79.3%), but Llama models remain highly susceptible (

<

14% RQ1) even when fine-tuned on their own failure cases. Together, these findings highlight substantial model-dependent limits of current robustness interventions and offer guidance for developing more trustworthy LLMs.

Fan Huang, Haewoon Kwak, Jisun An

Indiana University Bloomington, United States

LLM Evaluation Fine-Tuning LLM Reliability Belief Updating in LLMs Persuasion in Language Models Language Model Robustness

Jan 19, 2026

OI-Bench: An Option Injection Benchmark for Evaluating LLM Susceptibility to Directive Interference

Benchmarking large language models (LLMs) is critical for understanding their capabilities, limitations, and robustness. In addition to interface artifacts, prior studies have shown that LLM decisions can be influenced by directive signals such as social cues, framing, and instructions. In this work, we introduce option injection, a benchmarking approach that augments the multiple-choice question answering (MCQA) interface with an additional option containing a misleading directive, leveraging standardized choice structure and scalable evaluation. We construct OI-Bench, a benchmark of 3,000 questions spanning knowledge, reasoning, and commonsense tasks, with 16 directive types covering social compliance, bonus framing, threat framing, and instructional interference. This setting combines manipulation of the choice interface with directive-based interference, enabling systematic assessment of model susceptibility. We evaluate 12 LLMs to analyze attack success rates, behavioral responses, and further investigate mitigation strategies ranging from inference-time prompting to post-training alignment. Experimental results reveal substantial vulnerabilities and heterogeneous robustness across models. OI-Bench is expected to support more systematic evaluation of LLM robustness to directive interference within choice-based interfaces.

Yow-Fu Liou, Yu-Chien Tang, Yu-Hsiang Liu +1

Department of Computer Science, National Yang Ming Chiao Tung University, Taiwan

LLM Evaluation Adversarial Robustness of Language Models

Jan 14, 2026

Representational Similarity Alignment of Figurative Language in Humans and Large Language Models

Large language models (LLMs) often produce judgments similar to those of humans, but their interpretation of figurative language remains unclear. We compared human participants and four instruction-tuned LLMs (GPT-4, Gemma-2-9B, Llama-3.2, and Mistral-7B) on their ratings of 240 dialogue-based sentences spanning six linguistic traits: conventionality, sarcasm, funny, emotion, idiomacy, and Gen Z slang. While humans and LLMs were strongly aligned at the surface level, differences emerged in representational similarity patterns, particularly in idiomacy and Gen Z slang. GPT-4 showed the closest alignment to humans, though all models differed at the representational level, suggesting that similar outputs may arise from different underlying language representations. We also examined the relationship between model scale and representational similarity using the Gemma model series (Gemma-2-9B-IT, Gemma-3-27B-IT, and Gemma-4-31B-IT), and found that representational similarity patterns were largely unaffected by model size.

Samhita Bollepally, Aurora Sloman-Moll, Takashi Yamauchi +1

Figurative Language Understanding LLM Evaluation Pragmatic Reasoning in Language Models

Jan 13, 2026

QuantEval: A Benchmark for Financial Quantitative Tasks in Large Language Models

Large Language Models (LLMs) have shown strong capabilities in finance-oriented question answering, yet their evaluation on professional quantitative workflows remains fragmented and often stops at factual or document-grounded QA. We introduce QuantEval, a benchmark for evaluating whether LLMs can complete the compute--implement--validate loop required in quantitative finance. QuantEval contains 1,575 expert-validated instances across three complementary dimensions: (i) knowledge-based QA, (ii) quantitative mathematical reasoning, and (iii) quantitative strategy coding. The coding task is evaluated with a deterministic CTA-style backtesting framework that executes model-generated strategies and compares risk--return metrics against expert references, moving beyond syntax or template matching. We evaluate 13 open-source and proprietary LLMs, report prompt and reliability analyses, and observe substantial gaps relative to human experts, especially in multi-step reasoning and framework-constrained coding. We further conduct supervised fine-tuning and reinforcement learning experiments on domain-aligned data, showing that training can improve quantitative reasoning while also exposing trade-offs in factual recall. QuantEval releases the dataset, evaluation scripts, and full backtesting configuration to support reproducible research on LLMs for quantitative finance.

Zhaolu Kang, Junhao Gong, Wenqing Hu +13

Peking University · Tsinghua University · Fudan University +6

Financial Language Model Evaluation LLM Evaluation Numerical Reasoning in Language Models Quantitative Finance LLM Code Generation

Jan 13, 2026

How Order-Sensitive Are LLMs? OrderProbe for Deterministic Structural Reconstruction

Large language models (LLMs) excel at semantic understanding, yet their ability to reconstruct internal structure from scrambled inputs remains underexplored. Sentence-level restoration is difficult to evaluate automatically because scrambled sentences often admit multiple valid reorderings. We introduce OrderProbe, a deterministic benchmark for structural reconstruction using fixed four-character expressions in Chinese, Japanese, and Korean, which have a unique canonical order and thus support exact-match scoring. We further propose a diagnostic framework that evaluates models beyond recovery accuracy, including Semantic Accuracy, Logical Validity, Structural Consistency, Robustness, and Information Density. Experiments on twelve widely used LLMs show that structural reconstruction remains difficult even for frontier systems: zero-shot recovery frequently falls below 35%. We also observe a consistent gap between meaning-oriented generation and exact structural reconstruction, suggesting that structural robustness is not an automatic byproduct of semantic competence.

Zhaolu Kang, Yingjie He, Kehan Jiang +18

Hunyuan, Tencent · Peking University · Tsinghua University +8

Input Order Sensitivity in Language Models LLM Evaluation Language Model Robustness

Jan 13, 2026

STAGE: A Full-Screenplay Benchmark for Reasoning over Evolving Stories

Movie screenplays are rich long-form narratives that interleave complex character relationships, temporally ordered events, and dialogue-driven interactions. While prior benchmarks target individual subtasks such as question answering or dialogue generation, they rarely evaluate whether models can construct a coherent story world and use it consistently across multiple forms of reasoning and generation. We introduce STAGE (Screenplay Text, Agents, Graphs and Evaluation), a unified benchmark for narrative understanding over full-length movie screenplays. STAGE defines four tasks: knowledge graph construction, scene-level event summarization, long-context screenplay question answering, and in-script character role-playing, all grounded in a shared narrative world representation. The benchmark provides cleaned scripts, curated knowledge graphs, and event- and character-centric annotations for 150 films across English and Chinese, enabling holistic evaluation of models' abilities to build world representations, abstract and verify narrative events, reason over long narratives, and generate character-consistent responses.

Qiuyu Tian, Zequn Liu, Yiding Li +9

Southeast University, Nanjing, China · Beijing Zhongguancun Academy, Beijing, China · ZhuiWen Technology Co., Ltd., Beijing, China +1

KG Construction LLM Evaluation Long-Context QA

Jan 12, 2026

Calibration Is Not Enough: Evaluating Confidence Estimation Under Language Variations

Confidence estimation (CE) indicates how reliable the answers of large language models are and impacts user trust and decision-making. Existing evaluations mainly concern the alignment between confidence and correctness, but ignore the variability of language: confidence estimates should remain consistent under semantically equivalent prompts or answer variations, while changing when answer meaning differs, as this may indicate a change in correctness. Therefore, we introduce a novel evaluation framework based on three complementary properties: robustness to prompt perturbations, stability across semantically equivalent answers, and sensitivity to semantically different answers. We show that these metrics are largely independent from existing CE metrics, and that common CE methods often fail on them: while most methods achieve high robustness and stability, they struggle to distinguish semantically different answers, potentially because they do not effectively leverage generation-side information. Overall, our framework exposes overlooked limitations of current CE evaluations and provides guidance for selecting confidence estimators for real-world applications.

Yuxi Xia, Dennis Ulmer, Terra Blevins +3

Faculty of Computer Science, UniVie Doctoral School Computer Science · Faculty of Philological and Cultural Studies, University of Vienna, Austria · ILLC, University of Amsterdam, Netherlands +2

LLM Uncertainty Estimation Confidence Estimation in Language Models Language Model Self-Assessment LLM Evaluation Language Model Calibration LLM Reliability+1

Jan 12, 2026

Kinship Data Benchmark for Multi-hop Reasoning

Multi-hop kinship reasoning is a natural testbed for LLM compositionality, but existing benchmarks (notably CLUTRR) cover only the descriptive Eskimo system. We introduce KinshipQA, a procedurally-generated benchmark covering seven anthropologically-documented kinship systems (Eskimo, Sudanese, Hawaiian, Iroquois, Dravidian, Crow, Omaha) and up to six reasoning hops, with a tunable simulator horizon that eliminates exact-instance pretraining overlap. Evaluating six LLMs, we find a 40.9% accuracy drop when reasoning shifts from biological multi-hop to culturally-marked classification on the five non-descriptive systems. The drop holds for every non-descriptive system and is largest for the two skewing systems (Crow, Omaha), persists under chain-of-thought and few-shot prompting, and compounds with depth: at 5--6 hops cultural override falls to 10.6% while biological composition over the same chains remains at 58.6%. Under identical rule access humans reach 89.0% versus 50.7% for LLMs, so the questions are reliably solvable once the rule is supplied. Two follow-up experiments suggest distinct contributors. A fictional-rule control swapping system labels and kin terms for invented strings raises accuracy by 6.1%, implicating familiar English surface forms. An in-context-rule probe prepending the override rule helps skewing systems (+17.1%) but hurts non-skewing systems whose baseline already exceeds about 60% (-13.4%), consistent with a missing skewing prior alongside rule interference where the model already has a working approximation. Our code and data are publicly available on GitHub.

Tianda Sun, Dimitar Kazakov

Department of Computer Science University of York, UK

LLM Evaluation Synthetic Benchmark Generation Multi-Hop Reasoning

Jan 12, 2026

Stochastic CHAOS: Why Deterministic Inference Kills, and Distributional Variability Is the Heartbeat of Artifical Cognition

Deterministic inference is a comforting ideal in classical software: the same program on the same input should always produce the same output. As large language models move into real-world deployment, this ideal has been imported wholesale into inference stacks. Recent work from the Thinking Machines Lab has presented a detailed analysis of nondeterminism in LLM inference, showing how batch-invariant kernels and deterministic attention can enforce bitwise-identical outputs, positioning deterministic inference as a prerequisite for reproducibility and enterprise reliability. In this paper, we take the opposite stance. We argue that, for LLMs, deterministic inference kills. It kills the ability to model uncertainty, suppresses emergent abilities, collapses reasoning into a single brittle path, and weakens safety alignment by hiding tail risks. LLMs implement conditional distributions over outputs, not fixed functions. Collapsing these distributions to a single canonical completion may appear reassuring, but it systematically conceals properties central to artificial cognition. We instead advocate Stochastic CHAOS, treating distributional variability as a signal to be measured and controlled. Empirically, we show that deterministic inference is systematically misleading. Single-sample deterministic evaluation underestimates both capability and fragility, masking failure probability under paraphrases and noise. Phase-like transitions associated with emergent abilities disappear under greedy decoding. Multi-path reasoning degrades when forced onto deterministic backbones, reducing accuracy and diagnostic insight. Finally, deterministic evaluation underestimates safety risk by hiding rare but dangerous behaviors that appear only under multi-sample evaluation.

Tanmay Joshi, Shourya Aggarwal, Anusa Saha +7

Pragya Lab, BITS Pilani Goa, India · Raapid Lab, USA · Apple, USA +1

Emergent Abilities in Language Models LLM Reliability LLM Evaluation

Jan 12, 2026

Measuring Iterative Temporal Reasoning with Time Puzzles

Tool use, such as web search, has become a standard capability even in freely available large language models (LLMs). However, existing benchmarks evaluate temporal reasoning mainly in static, non-tool-using settings, which poorly reflect how LLMs perform temporal reasoning in practice. We introduce Time Puzzles, a constraint-based date inference task for evaluating iterative temporal reasoning with tools. Each puzzle combines factual temporal anchors with (cross-cultural) calendar relations and may admit one or multiple valid dates. The puzzles are algorithmically generated, enabling controlled and continual evaluation. Across 13 LLMs, even the best model (GPT-5) achieves only 55.3% accuracy without tools, despite using easily searchable facts. While web search improves performance, models perform substantially better when constraints are rewritten with explicit dates, removing the need for factual lookup. These results reveal a gap in reliable tool use for iterative temporal reasoning.

Zhengxiang Wang, Zeyu Dong

Department of Applied Math and Statistics Stony Brook University

LLM Evaluation Temporal Reasoning in Language Models Temporal Reasoning LLM Tool Use

Jan 10, 2026

BizFinBench.v2: Towards Reliable LLMs in Finance via Real-User Data and Offline/Online Bilingual Evaluation

Large language models are becoming increasingly significant in financial applications. Nevertheless, prevailing benchmarks are largely dependent on simulated or generic data, which leads to a significant gap between reported performance and actual efficacy in real-world scenarios. To tackle this challenge, we present BizFinBench.v2, the first integrated offline and online benchmark built upon authentic user query-response data from both Chinese and U.S. equity markets. It comprises 28,860 questions across eight offline and two online tasks. Experimental results show that GPT-5 achieves a mere 61.5% accuracy, still failing to meet the practical business requirement (84.8%). Among the evaluated commercial models, DeepSeek-R1 exhibits superior investment efficacy. Error analysis grounded in real financial practice reveals persistent limitations in existing models. By overcoming the constraints of prior benchmarks, BizFinBench.v2 provides a substantiated foundation for advancing LLM deployment in the financial sector. Our data and code are available at https://github.com/HiThink-Research/BizFinBench.v2.

Xin Guo, Rongjunchen Zhang, Guilong Lu +4

HiThink Research · Shanghai University of Finance and Economics

Financial Services Multilingual Language Model Evaluation LLM Evaluation Finance Financial Reasoning Benchmarks