Unreliable Reasoning

Traditionally, the reliability of news publishers is assessed by expert organisations that evaluate editorial practices, transparency and factual standards at source. When this process is translated into a computational approach, the problem is often formulated at the level of individual articles, with models being trained on a set of pre-labelled articles and their performance being evaluated in a test phase. In this work, we investigate news source reliability inference as a source-level prediction problem. We propose a two-stage framework in which transformer-based language models first estimate the reliability of individual articles and subsequently aggregate article-level predictions to infer the reliability of previously unseen publishers. To approximate realistic deployment conditions, we enforce a strict publisher-disjoint evaluation protocol, ensuring that no publisher appears in both training and test sets. Experiments on 19,476 political news articles from 439 English-language publishers labeled with NewsGuard reliability ratings show that aggregation substantially improves robustness and performance, increasing accuracy from approximately 0.60 at the article level to 0.69 at the publisher level. Finally, we analyze how prediction errors vary across political orientations, revealing statistically significant associations between political leaning and misclassification patterns. Overall, our findings show that publisher reliability can be inferred from aggregated textual signals alone, supporting scalable and content-based approaches to automated news source assessment.

In this paper, we present TRACE (Tokenized Robust Attention for Contact-Aware Estimation), an end-to-end learned proprioceptive odometry estimator for legged robots under unreliable contact conditions. The proposed estimator directly predicts relative displacement, relative rotation, and body-frame velocity from a recent history of onboard inertial and joint measurements. To improve robustness under unreliable contact conditions, we introduce a foot-aware cross-attention module that adaptively weights IMU and leg-wise kinematic tokens without relying on manually defined contact or slip thresholds. The estimator is trained with direct supervision and two physics-inspired auxiliary losses that promote kinematic consistency and reliable use of leg information. To reduce policy-specific overfitting and consequently improve sim-to-real transfer, simulation training incorporates policy randomization, followed by partial real-world fine-tuning of the temporal encoder and prediction head. Experiments across diverse indoor and outdoor terrains demonstrate consistent reductions in position drift compared with classical filtering-based, hybrid, and purely learning-based baselines. Ablation studies further validate the contributions of the proposed training objectives, policy randomization, and real-world fine-tuning, particularly under unreliable contacts and sim-to-real mismatch.

Black-box language-model reliability is commonly pursued by sampling, prompting, voting, verifying, or iteratively revising individual answers. We ask a prior question: \emph{what determines whether a collection of black-box responses is recoverable at all?} We represent responses to typed transformations of a query as a \emph{relational response field} (RRF). Edge transports encode how valid responses must change under paraphrase, scaling, decomposition, refactoring, or other task symmetries; anchors encode independently trusted evidence such as execution or a verifier. For relation operator $D$, anchor operator $A$, and at most $k$ corrupted response nodes, we identify $γ_k(D,A)$ as the intrinsic difficulty of black-box response recovery. It is positive exactly when every $k$-node corruption is identifiable; it gives a deterministic stability bound proportional to $1/γ_k$; and a matching two-point minimax lower bound shows that no estimator can improve this dependence. Thus consistency is not truth: relation-only methods are blind to null directions, including shared hallucinations. We derive sparse field-repair algorithms while separating information-theoretic identifiability from the stronger null-space conditions required by convex optimization. Controlled theorem tests and black-box mathematics/code experiments evaluate four theory-fixed consequences: consistency--truth separation, anchor phase transitions, redundancy saturation, and cross-model, cross-task prediction of repair difficulty. The results support $γ_k(D,A)$ as a measurable property of a response-recovery instance, rather than a score attached to one repair heuristic.

Content-based detection of unreliable news is increasingly difficult, as low-reliability sources mimic credible journalism and generative AI makes fabricated content harder to flag. We ask whether network structure can improve news reliability classification, taking a domain-level approach that shifts the focus from individual articles to source reliability. From URL-sharing patterns in Telegram chats, we build a statistically validated domain co-sharing network and find assortative mixing by reliability: low-reliability domains group together, as do reliable ones. Exploiting this structure, we compare Graph Neural Networks against network-unaware baselines using both content-aware features (multilingual text embeddings) and content-agnostic features (spreading dynamics). GNNs consistently outperform Multi-Layer Perceptrons on identical features, with GraphSAGE best in both settings (accuracy 0.63 with content, 0.53 without), a 13-14% relative gain over the network-unaware baseline. Network topology thus systematically improves domain reliability assessment, and remains effective even when content analysis is infeasible.

Communication allows large language model agents to pool evidence, but it also creates paths along which an erroneous claim can spread. We formulate a correction-aware network model that tracks susceptible, exposed, infectious, and corrected agents and derive its early-invasion condition for heterogeneous communication networks. We then couple this propagation model to an analytic majority-vote benchmark in which a clean-task reliability target imposes a minimum connectivity requirement. Under fixed exposure per communication edge, reliability and error control impose opposing graph constraints. We characterize when their intersection is empty and when it contains an intermediate connectivity range, and identify regular graphs that attain the smallest invasion factor within the reliable graph class when such graphs exist. Under a fixed sender budget, the homogeneous first-order threshold is independent of network density, showing that the communication-budget convention determines whether added edges increase early propagation risk. Finite-network simulations on 21,000 trajectories illustrate these directional predictions. A controlled grok-4.3 experiment then evaluates three six-node topologies on 36 new closed-world tasks, with a balanced 12-task subset continued to full cascades. Mean first-generation offspring increased from 0.667 to 1.333 and 1.667 as degree increased from 2 to 4 and 5, while the adoption fraction among exposed neighbours remained 0.333. Mean non-seed erroneous adoption in the full-cascade subset was 0.200, 0.333, and 0.333. Together, these results provide a tractable basis for selecting connectivity under explicit reliability and propagation constraints.

Large language models can arrive at the same answer through reasoning paths that are unstable, contradictory, or difficult to rank consistently -- a failure mode especially prevalent in multi-step deductive reasoning. Existing methods assess reliability primarily through output dispersion -- measuring how much sampled answers differ -- but this discards a complementary signal: whether the model can consistently rank competing reasoning candidates. We propose structural uncertainty, a consistency-aware framework derived from the stability of self-preference-induced rankings over sampled reasoning solutions. Given a query, we generate multiple candidate solutions and ask the model to judge pairwise preferences among its own outputs. We aggregate self-preferences into ranking distributions via Bradley-Terry modeling with PageRank, and decompose the signal into two entropy-based components: across-trial ranking instability and within-trial candidate ambiguity. Across five LLMs and eight benchmarks, structural signals provide information complementary to answer dispersion: on logical and mathematical reasoning tasks, the combination improves identification of unreliable instances, while on factual retrieval the structural signal collapses toward uniformity, diagnosing a regime boundary where reasoning-level consistency evaluation is uninformative. The two components relate differently to accuracy: within-trial ambiguity correlates positively with correctness -- consistent with settings where multiple plausible solution paths remain competitive -- while across-trial instability correlates negatively, signaling unreliable reasoning. Structural uncertainty is best understood not as a universal confidence estimator, but as a regime-sensitive evaluator of logical reasoning consistency.

Multi-agent systems, in which multiple large language model agents solve problems through turn-based interaction, are increasingly deployed in high-stakes settings such as medical diagnosis, legal analysis, and forensic decision-making. Their reliability can be at risk when single agents reason from incorrect or misleading context, e.g., from tool calls, since errors may propagate through agent interactions. This work studies this risk by injecting intent-based misinformation into benign single-agent and multi-agent systems across reasoning, knowledge, and alignment tasks. We find that misinformation can degrade single-agent performance and persists across multi-agent debate, with agents often retaining answers introduced by misinformed peers. Nevertheless, multi-agent debate reduces the resulting performance degradation compared to single-agent prompting, especially when most agents are not exposed to misinformation. Robustness depends on group composition and decision protocol. Consensus can be more stable than voting under peer pressure, while majorities can often steer misinformed agents back toward correct answers. Our results show that misinformation robustness in multi-agent systems depends on the underlying model and also on how agents exchange information and aggregate decisions.

We ask a structural question: given unreliable elementary problem-solvers, what organizations of them solve hard problems reliably, and what are the limits? We develop a $decomposition~algebra$: elementary solvers are morphisms in a stochastic category, and four combinators (sequential composition, parallel ensembling, verification gating, and recursive reduction) generate the space of compound solvers. We equip this algebra with two homomorphisms, a $reliability$ valuation into the ordered monoid $([0,1],\le)$ and a $cost$ valuation into a commutative semiring, and we derive the composition laws that govern how reliability flows through structure. Our central results are (i) a $verification~odds~law$ (the result that names this report), showing that a verification gate multiplies the odds of correctness by the verifier's likelihood ratio $Λ$, so that $k$ conditionally independent gates yield geometric amplification; (ii) a $reliability~amplification~theorem$, giving target reliability $1-δ$ at $O(\log 1/δ)$ verification depth whenever $Λ>1$; and (iii) a $threshold~dichotomy$: above the critical parameters reliability can be driven arbitrarily close to one at logarithmic cost, while at or below them no amplification is possible. We then show that $self-organization$ is the least fixed point of a monotone improvement operator on the complete lattice of strategies, and that this fixed point equalizes marginal log-odds gain per unit cost. Finally, we prove matching limits: an information ceiling bounds per-gate amplification by a divergence quantity; shared error causes create a strictly positive voting floor, so diversity is $necessary$ for unbounded amplification. Reliability, in short, is neither free nor magical: it is bought with independent information, arranged by composition, and bounded by the verifier.

Retrieval-Augmented Generation (RAG) is widely used to improve the factual reliability of large language models (LLMs) by grounding answers in retrieved evidence. In misinformation-rich environments, however, retrieved content may include plausible but incorrect information, raising concerns about the reliability of RAG-based information access systems. In this work, we propose an evaluation protocol to systematically test how the RAG system handles conflicts between parametric knowledge and evidence retrieved from context with varying amounts of misleading information. We target correct answers to factoid questions that the model responds to correctly, even when there is no retrieval, and use this to test the system with clean, poisoned, and mixed evidence. The proposed analytical framework combines parametric override and confidence metrics to assess when and how misleading information affects the generation process of LLMs. This study aims to provide insights into the robustness of RAG systems in information disorder scenarios.

Improving the quality of model-generated summaries, especially factuality, the accuracy of a summary with respect to its source content, remains a challenge. While reranking could select the optimal output from multiple generated candidates, it is limited to only using the source as guidance, resulting in unreliable summaries. To address this limitation, we propose ConSUM that reranks candidate summaries by considering two factors: consistency to the source document and consensus among the other candidates. Consensus is established using Minimum Bayes Risk (MBR) decoding over the set of generated summaries, while ensuring consistency by employing factuality-aware metrics that compare the summary against the source. Rigorous testing demonstrates that our system is competitive with existing methods, with human evaluations further confirming that its generated summaries are preferred over those from other systems. Our code is available at https://github.com/naist-nlp/ConSUM .

Persuasive dialogue requires reasoning about others' latent mental states, a capability known as Theory of Mind (ToM). However, due to reliance on simple prompting strategies and insufficient ToM knowledge, existing LLMs often fail to capture the intrinsic dependencies among mental states, leading to fragmented representations and unstable reasoning. To address these challenges, we introduce the ToM-based Persuasive Dialogue (ToM-PD) task, grounded in the Belief-Desire-Intention (BDI) framework, which explicitly models the sequential dependencies among mental states in multi-turn dialogues. To facilitate research on this task, we construct a large-scale annotated dataset, ToM-based Broad Persuasive Dialogues (ToM-BPD), capturing fine-grained mental states and corresponding persuasive strategies. We further propose Think Thrice Before You Speak (TTBYS), a knowledge-enhanced stepwise reasoning framework that leverages both explicit and implicit prior experiences to improve LLMs' inference of desires, beliefs, and persuasive strategies. Experimental results demonstrate that Qwen3-8B equipped with TTBYS outperforms GPT-5 by 1.20%, 22.80%, and 16.97% in predicting desires, beliefs, and persuasive strategies, respectively. Case studies further show that our approach enhances interpretability and consistency in reasoning.

This paper studies the Byzantine Agreement problem where the nodes have access to a predictor that flags nodes for suspicion of faulty (Byzantine) behavior. We focus on algorithmic resilience -- the maximum number of faulty nodes an algorithm can tolerate -- and present algorithms and impossibility results whose resilience depend on the accuracy of the predictor. As our first main result, we bring a complete characterization of the consistency--robustness trade-offs in both the non-authenticated and authenticated settings: for $n$ nodes and a parameter $α\in [0, 1]$, we present algorithms that tolerate up to $α\cdot n$ faulty nodes when the predictor is correct (consistency), and up to $\frac{1-α}{2} \cdot n - 1$ faulty nodes when the predictor is arbitrarily wrong (robustness); in the authenticated setting the robustness bound improves to $(1-α) \cdot n - 1$. These trade-offs are exactly tight as we show that one additional faulty node renders the problem impossible. Our second main result characterizes smoothness: the rate at which resilience degrades as the predictor becomes less accurate. We show that resilience linearly decreases in the number of wrong predictions as long as that number stays within a constant fraction of $n$. Concretely, in the non-authenticated setting each additional wrong prediction loses one unit of resilience, whereas in the authenticated setting the decline is halved since two wrong predictions are needed to lose one unit of resilience.

While reasoning has become a central capability of large language models (LLMs), the reasoning patterns required for different scenarios are often misaligned. Mathematical reasoning typically relies on intrinsic logic to solve closed-world problems in a single response, whereas agentic reasoning requires not only internal reasoning but also multi-turn interaction with external environments, interleaving thought and action. This misalignment prevents mathematical and agentic reasoning from effectively benefiting from each other, often yielding unstable reasoning behavior and only limited performance gains under multi-task learning. In this paper, we propose M2A, a novel paradigm that synergizes mathematical and agentic reasoning via model merging. To avoid overfitting to superficial reasoning patterns under joint training, M2A operates directly in parameter space: it identifies the feature subspace critical for agent behavior, and merges the mathematical reasoning task vector only along its null space, thereby injecting reasoning capability along directions that do not perturb agent behavior. Unlike SFT or RL, M2A requires no additional gradient-update and exposes the merging coefficient as a simple knob for controlling reasoning length. Experiments in a challenging real-world coding agent setting show that our method effectively extends agentic reasoning depth and delivers substantial performance improvements. Applied to a fine-tuned Qwen3-8B, M2A improves its SWE-Bench Verified resolved rate from 44.0% to 51.2% without retraining the model. Code is available at https://github.com/laplucky/M2A.git.

Large language model (LLM) agents increasingly collaborate over peer-to-peer networks to improve their reliability. However, these same interactions can also become a source of vulnerability, as unreliable or Byzantine agents may sway neighboring agents toward incorrect conclusions and degrade overall system performance. Existing methods rely on leader-based coordination or self-reported confidence, both of which are susceptible to adversarial manipulation. We study decentralized LLM multi-agent systems (LLM-MAS) and propose Self-Anchored Consensus (SAC), a fully decentralized iterative filter-and-refine protocol in which agents iteratively exchange responses, locally evaluate and filter unreliable messages, and refine their own outputs. We present $(F{+}1)$-robustness conditions for the communication graph that ensure honest agents preserve and propagate reliable information despite Byzantine influence. Experiments on mathematical and commonsense reasoning benchmarks show that SAC effectively suppresses Byzantine influence and consistently improves performance across diverse communication topologies, whereas prior methods degrade under adversarial conditions.

Collaborative machine learning (CML) enables multiple clients to train a global model jointly in a data-distributed setting. To address data privacy and communication efficiency, one-shot CML has been increasingly adopted, where clients communicate with the server only once by sharing synthetic or processed proxy data. This single-round communication, however, eliminates the possibility of iterative correction at the server, making the learning process particularly vulnerable to client unreliability. In this setting, unreliable clients, whether malicious or non-malicious, may provide biased proxy data that favors certain groups, thereby degrading the fairness of the global model and harming minority or unprivileged groups. In this work, we propose a server-side defense framework based on a bilevel optimization formulation. The proposed approach learns client-level weights to mitigate the influence of biased client proxy data while enforcing fairness constraints by using a very small trusted root dataset available at the server. Experimental results on benchmark datasets show that our method improves fairness with little accuracy loss under biased proxy data contributions from unreliable clients. Moreover, the proposed approach remains effective even when unreliable clients make up a majority of the system, consistently outperforming other existing methods.

As artificial intelligence (AI), including machine learning (ML) models and foundation models (FMs), are increasingly deployed in high-stakes domains, ensuring their trustworthiness has become a central challenge. However, the core trustworthy AI objectives, such as fairness, robustness, privacy, and explainability, are hard to achieve simultaneously, especially while preserving utility. This position paper argues that causality is necessary to understand and balance trade-offs in performance and multiple objectives of trustworthy AI. We ground our arguments in re-interpreting trustworthy AI trade-offs as incompatible invariance requirements under different changes to the data-generating process. We then illustrate this argument through case-study analyses from the literature and a stylized synthetic-data simulation, showing that causality provides a unifying framework for understanding how trade-offs in trustworthy AI arise and how they can be softened or resolved through selective invariance. This perspective applies to both classical ML models and large-scale FMs. Finally, we outline open challenges and opportunities for using causality to build both trustworthy and high-performing AI.

Large language models are widely adopted as automated evaluation judges, yet the stability of their verdicts under semantically equivalent prompt rephrasings remains largely unexamined. We conduct a systematic empirical study of prompt-induced decision instability across multiple evaluation tasks and judge architectures. To facilitate this analysis, we release JudgeSense, a benchmark comprising hand-validated prompt-paraphrase pairs spanning factuality, coherence, relevance, and preference, drawn from established NLP benchmarks and accompanied by comprehensive decision logs. The benchmark enables the measurement of judge stability across equivalent prompts, allowing researchers to assess whether stability correlates with model scale or instruction-tuning, and to identify which tasks are most sensitive to prompt wording. Our evaluation reveals that coherence remains the primary task for distinguishing judge behavior, while factuality judgments demonstrate high stability under standard conditions. Pairwise evaluation tasks consistently exhibit position bias. Crucially, we find that model scale is not a reliable proxy for consistency; notably, as an interesting result in our analysis, the largest and newest models are not the most consistent.

Repair, an important resource for resolving trouble in human-human conversation, remains underexplored in human-LLM interaction. In this study, we investigate how LLMs engage in the interactive process of repair in multi-turn dialogues around solvable and unsolvable math questions. We examine whether models initiate repair themselves and how they respond to user-initiated repair. Our results show strong differences across models: reactions range from being almost completely resistant to (appropriate) repair attempts to being highly susceptible and easily manipulated. We further demonstrate that once conversations extend beyond a single turn, model behavior becomes more distinctive and less predictable across systems. Overall, our findings indicate that each tested LLM exhibits its own characteristic form of unreliability in the context of repair.

Computer-use agents have rapidly improved on real-world tasks such as web navigation, desktop automation, and software interaction, in some cases surpassing human performance. Yet even when the task and model are unchanged, an agent that succeeds once may fail on a repeated execution of the same task. This raises a fundamental question: if an agent can succeed at a task once, what prevents it from doing so reliably? In this work, we study the sources of unreliability in computer-use agents through three factors: stochasticity during execution, ambiguity in task specification, and variability in agent behavior. We analyze these factors on OSWorld using repeated executions of the same task together with paired statistical tests that capture task-level changes across settings. Our analysis shows that reliability depends on both how tasks are specified and how agent behavior varies across executions. These findings suggest the need to evaluate agents under repeated execution, to allow agents to resolve task ambiguity through interaction, and to favor strategies that remain stable across runs.

Conventional test-time adaptation (TTA) approaches typically adapt the model using only a small fraction of test samples, often those with low-entropy predictions, thereby failing to fully leverage the available information in the test distribution. This paper introduces DualTTA, a novel framework that improves performance under distribution shifts by utilizing a larger and more diverse set of test samples. DualTTA identifies two distinct groups: one where the model's predictions are likely consistent with the underlying semantics, and another where predictions are likely incorrect. For the first group, it minimizes prediction entropy to reinforce reliable decisions; for the second, it maximizes entropy to suppress overconfident errors and unlearn spurious behavior. These groups are adaptively selected using a new reliability criterion that measures prediction stability under both semantic-preserving and semantic-altering transformations, addressing the limitations of purely entropy-based selection. We further provide theoretical analysis and empirical justification showing that our approach enables a tighter separation between reliable and unreliable samples, in the context of their suitability for adaptation, leading to provably more effective model updates.

LLM-driven multi-agent frameworks address complex reasoning tasks through multi-role collaboration. However, existing approaches often suffer from reasoning instability, where individual agent errors are amplified through collaboration, undermining overall performance. Current research mainly focuses on enhancing high-capability agents or suppressing unreliable outputs to improve framework effectiveness, while systematic identification and reinforcement of performance-limiting agents receive less attention. To address this gap, we propose WORC, a \underline{w}eak-link \underline{o}ptimization framework for multi-agent \underline{r}easoning and \underline{c}ollaboration, grounded in the weak-link principle. WORC follows a two-stage workflow. In the weak agent localization stage, task features are constructed, and a meta-learning-based weight predictor trained on optimal configurations identified by swarm intelligence algorithms (SIAs) enables zero-shot mapping from these features to agent performance weights, where the agent with the lowest predicted weight is identified as the weak agent. In the weak-link optimization stage, an uncertainty-driven allocation strategy assigns additional reasoning budgets to weak agents, with lower predicted weights leading to larger repeated-sampling quotas to compensate for reliability deficiencies. Experimental results show that WORC achieves an average accuracy of 82.2% on reasoning benchmarks while improving framework stability and cross-architecture generalization, suggesting that compensating for weak links, rather than reinforcing strengths alone, enhances the robustness of multi-agent systems.

Large Language Models (LLMs) are poised to disrupt knowledge work, with the emergence of delegated work as a new interaction paradigm (e.g., vibe coding). Delegation requires trust - the expectation that the LLM will faithfully execute the task without introducing errors into documents. We introduce DELEGATE-52 to study the readiness of AI systems in delegated workflows. DELEGATE-52 simulates long delegated workflows that require in-depth document editing across 52 professional domains, such as coding, crystallography, and music notation. Our large-scale experiment with 19 LLMs reveals that current models degrade documents during delegation: even frontier models (Gemini 3.1 Pro, Claude 4.6 Opus, GPT 5.4) corrupt an average of 25% of document content by the end of long workflows, with other models failing more severely. Additional experiments reveal that agentic tool use does not improve performance on DELEGATE-52, and that degradation severity is exacerbated by document size, length of interaction, or presence of distractor files. Our analysis shows that current LLMs are unreliable delegates: they introduce sparse but severe errors that silently corrupt documents, compounding over long interaction.

Large language models are increasingly integrated into decision-making in areas such as healthcare, law, finance, engineering, and government. Yet they share a critical limitation: they produce fluent outputs even when their internal reasoning has drifted. A confident answer can conceal uncertainty, speculation, or inconsistency, and small changes in phrasing can lead to different conclusions. This makes LLMs useful assistants but unreliable partners in high-stakes contexts. Humans exhibit a similar weakness, often mistaking fluency for reliability. When a model responds smoothly, users tend to trust it, even when both model and user are drifting together. This paper is the first in a five-paper research series on stabilising human-AI reasoning. The series proposes a two-layer approach: Parts II-IV introduce human-side mechanisms such as uncertainty cues, conflict surfacing, and auditable reasoning traces, while Part V develops a model-side Epistemic Control Loop (ECL) that detects instability and modulates generation accordingly. Together, these layers form a missing operational substrate for governance by increasing signal-to-noise at the point of use. Stabilising interaction makes uncertainty and drift visible before enforcement is applied, enabling more precise capability governance. This aligns with emerging compliance expectations, including the EU AI Act and ISO/IEC 42001, by making reasoning processes traceable under real conditions of use. The central claim is that fluency is not reliability. Without structures that stabilise both human and model reasoning, AI cannot be trusted or governed where it matters most.

Fairness audits are a key component of responsible machine-learning deployment. Yet, audit-recommendation reliability under incomplete protected-label access is still poorly understood. In this work, we focused on protected-label missingness in fairness mitigation audits. We introduced a seed-calibrated stress test to separate missingness effects from seed-to-seed movement already present under complete labels. Across ACS/Folktables tasks, missingness settings that retain some protected labels usually do not move selected mitigation methods beyond a complete-label seed-to-seed baseline. At $0$ protected-label access, candidates collapse to an empirical-risk-minimization baseline and deterministic tie-breaking rather than revealing a broad missingness effect. We also found that threshold optimization can turn fairness gains on a single protected axis into intersectional harm above a seed baseline, and this threshold-optimizer finding persists under random-forest validation. Overall, our results highlight that protected-label missingness should be reported with seed-null calibration, candidate-set context, and intersectional consequences before it is treated as evidence of audit fragility.