CS Paper Radar

Same-Number Citation Swaps: Stress-Testing Jev as a Financial Evidence Judge

Chuhong Xu, Bo Su, Ziyao Chen, Ruiyang Xu, Shimeng Dai, Xinyu Qiu

cs.CL ·submitted 2026-10-06 16:57 UTC · v2
published

Abstract

Financial reports repeat values across periods, metrics and accounting lines, allowing an LLM-generated calculation to be numerically correct while citing the wrong financial role. We evaluate what probabilistic evidence verification adds beyond number matching using Jev as a source-support verifier for GPT-4.1-mini calculation traces. A signed-number-at-pointer baseline explains most recovery over exact quotation checks. To isolate the remaining role-recognition problem, we hold operands and arithmetic fixed, move citations between same-number cells, and retain controls that express equivalent facts. These contrasts reveal both wrong-role citations that pass and valid alternative citations that are withheld. Explicit column labels improve selected wrong-role decisions while also lowering support for some equivalent evidence. A constructed follow-up on 36 new source pages, labeled by a non-author reviewer, extends this evaluation and exposes the same tradeoff between detecting role errors and retaining valid citations. The contribution is a controlled evaluation that identifies what a probabilistic financial verifier distinguishes when numerical matching is held fixed. For LLM-based financial assistants, it makes numerical correctness, cited-role support and acceptance outcomes separately assessable.

Comment
counterfactual citation perturbation, evidence attribution verification, financial document question answering, Jev, LLM-as-a-judge, probabilistic source verification, tabular numerical reasoning. Accepted for presentation at the Agenthon workshop at NeurIPS 2026

Closest papers

Nearest neighbours by meaning across the whole index.

A Citation-Grounded Benchmark for Trustworthy Earnings Call Transcript Analysis with Large Language Models

cs.CL cs.AI ·2026-10-01 ·Yingzhu Zhao, Vlad Pandelea, Han Yuan, Bo Hu, Wuqiong Luo +2 more
benchmark similarity 0.63
Abstract

Large language models (LLMs) have been increasingly used for financial document analysis, including earnings call transcripts (ECTs). Beyond generating standalone claims, users increasingly prefer grounded analyses that pair claims with verifiable citations from source documents to enable independent validation. However, evaluating such analytical claims typically requires extensive expert annotation, which is costly and difficult to scale, and real-world financial analysis commonly involves long context-question-answer triplets, further increasing task complexity. To address these challenges and benchmark the current landscape of grounded analysis by LLMs, we propose a numeric evidence evaluation method that enables groundedness assessment without reliance on expert annotation. We also introduce an automated dataset construction pipeline and construct ECTs-100 from the top 100 constituents of the S&P 500 to support benchmark of both groundedness and correctness. In addition, we examine conscious incompetence, a practical failure mode in financial analysis in which LLMs must detect when available evidence is insufficient and refrain from producing unsupported hallucinations. Empirical results show that LLMs perform well in groundedness but face notable limitations in correctness, with informational insufficiency presenting an additional challenge.

0.17

Debiasing as a Measurement Intervention: Calibrated Ties and Resolution Loss in LLM-as-a-Judge Evaluation

cs.DL cs.AI ·2026-09-11 ·Liang Zhao, Yong Wang, Jiangzhe Chen
top venuebenchmark similarity 0.62
Abstract · Accepted at Emnlp 2026

LLM-as-a-judge protocols are commonly debiased by instructing judges to ignore presentation cues such as citation formatting, source labels, and evidence-display style. We show that this intervention can suppress bias while damaging the resolution of the measurement instrument. We introduce TraceJudgeBench, a diagnostic benchmark for auditing citation-like artifacts in RAG and agent-workflow evaluation, covering content-equivalent pairs, citation ablations, correctness conflicts, human-validated soft and moderate quality gaps, prompt-strength ladders, decoupled judging, and a controlled workflow-ranking probe. Across GPT-5.5, Claude Sonnet 4.6, and DeepSeek V4-Flash, stronger anti-citation prompts reduce worse-cited wins from up to 50.5% to 0%; yet some operating points already convert validated moderate-gap decisions into Tie before the strict stress-test endpoint, while correctness-conflict accuracy remains at or above 93.0%. A second, 50-pair FinQA moderate-gap construction reproduces the qualitative frontier, and open-weight Qwen2.5-14B-Instruct-AWQ and Gemma-3-12B-IT runs reproduce the central HotpotQA frontier. TRACE-style decoupling recovers 96.5-100.0% better-plain resolution across the reported settings. Human validation separates three meanings of Tie: correct equivalence Tie, calibrated soft-boundary Tie, and resolution-destroying Tie on validated quality gaps. We frame debiasing as a measurement intervention whose bias suppression, resolution retention, Tie cost, and protocol cost must be reported jointly. The supplementary artifact contains benchmark splits, prompts, raw judge outputs, validation summaries, and analysis.

0.37

CITECHOICE: A Causal Audit of How Document Presentation Redistributes Citation Credit in Agentic Search

cs.CL ·2026-09-14 ·Sriram Selvam, Anneswa Ghosh
similarity 0.60
Abstract

When several retrieved sources support the same claim, an answer engine cites some but not others. We call this decision citation allocation and introduce CITECHOICE, a causal audit of authentic multi-turn agentic search. From 129 everyday-query transcripts, CITECHOICE selects 113 same-call document pairs with independently verified support for the same pre-specified fact, without observing ranks or answer outcomes; blinded human review confirms 103. It runs a hash-verified 2-by-2 replay crossing pair order with jointly generated, fidelity-checked structured and prose renderings of one target while the rest of the transcript remains fixed. Three results emerge. First, and most importantly, structured rendering concentrates citation credit rather than clearly increasing source admission. It raises target citation count by +0.50 citations per answer (95 percent CI [+0.20, +0.84]; Holm-adjusted p=.033), without increasing total citations or reducing competitor credit. The pre-specified incidence effect (whether the target is cited at all) is +4.5 percentage points and inconclusive (95 percent CI [-1.4, +10.4]; p=.168). Second, observational position differences exceed controlled reordering effects: the citation-rate gap between rank 1 and rank 5 is 42.3 percentage points, compared with +7.9 percentage points in the main replay and 0.0 percentage points held out. Third, citation evaluation has a measurable noise floor. Although the aggregate count effect repeats under fresh decoding of 30 frozen families, 15 percent of binary decisions change and decoding accounts for an estimated 45 percent of single-generation family-effect variance. Together, these findings isolate what survives control: presentation can causally redistribute visible citation credit within frozen transcripts. They do not establish reliable source admission, a pure formatting mechanism, or a general rank advantage.

0.18

NumTrace: Auditing the Number Provenance of Financial Analyst Agents

cs.CL cs.AI ·2026-10-26 ·Tianlei Zhu, Matt White, Xue Liu, Xiao-Yang Liu
benchmark similarity 0.59
Abstract · 22 pages, 8 figures, 12 tables

A numerical claim generated by a financial analyst agent can match a number in a company filing but refer to the wrong entity, concept, period, unit or scope. A check based on reference answers can mark such an answer as wrong but does not explain why it fails. In this paper, we present NumTrace, a new benchmark that audits the provenance of each numerical claim made by financial analyst agents: its source and financial semantics. NumTrace currently has 540 questions over 2,163 SEC EDGAR documents. To the best of our knowledge, it is the first numerical benchmark about who holds a company's shares (Form 13F) and what its executives are owed (Item 402(i)). On a frozen snapshot of these documents, we build the reference answers according to accounting rules. NumTrace then gives each claim three separate verdicts: value, binding and support. By retaining separate value and binding verdicts, NumTrace detects right number, wrong account (RNWA) errors: a claim has an RNWA error when it passes the value check but fails the binding check. The RNWA rate quantifies the gap between value-only and joint value-and-binding pass rates. Finally, we audit three agents, Claude Opus 5.5, GPT-6 Astra and DeepSeek-V4.1-Flash. RNWA errors make up 67.5-81.8% of each agent's failing claims. In particular, on 13F co-filing group questions, about 91% of claims pass the value check, but fewer than half pass the binding check.

0.57

FinVerify: A Verification Framework for Financial Agents

cs.AI ·2026-11-16 ·Mao Guan, Qian Chen, Ge Meng, Xiao-Yang Liu
published similarity 0.59
Abstract · 15 pages, 2 figures. Accepted to the NeurIPS 2026 Workshop on Who Verifies the Agents? Toward Reliable Agent Development

Financial agents produce numerical claims that can be wrong even when their calculations execute correctly, due to mismatched reporting periods, unsupported sources, incompatible units, or incorrect financial definitions. We introduce FinVerify, an evidence-grounded framework for verifying financial agents' numerical claims. Based on an analysis of 190 incorrect trajectories from the Agent Finance Reasoning dataset, we define five error categories--temporal integrity, source fidelity, comparability, accounting integrity, and financial-model fidelity--and operationalize them as verification protocols. An LLM structures claims and supporting evidence, while deterministic validators check financial requirements; when authoritative formulas are available, Z3 checks consistency with grounded inputs. FinVerify accepts only if all applicable checks pass, rejects proven violations, and abstains when evidence is insufficient. We evaluate the framework against a Direct baseline and an LLM judge across five fixed-candidate benchmarks. On Agent Finance Reasoning, accepted-claim precision increases from 56.4% for the LLM judge to 76.7% at 81.2% coverage. On XBRLFiling, FinVerify achieves 99.8% accepted-claim precision and 97.0% decision accuracy; on FinanceBench-67, it reaches 87.9% accepted-claim precision versus 69.7% for the judge. Improvements are smaller on FinQA and TAT-QA, where authoritative evidence is limited. These results suggest that explicit, evidence-backed verification can improve financial-agent reliability while exposing trade-offs between correctness, false rejection, and abstention.

0.34

AtomCite: Verification and Correction of Supplied Page-Level Citations in Multi-Page Documents

cs.CL cs.AIcs.IR ·2026-09-05 ·Chen Qian, Yimeng Wang, Yu Chen, Lingfei Wu, Andreas Stathopoulos
benchmark similarity 0.57
Abstract

Large language models answering questions over multi-page documents are expected to cite the supporting pages, yet supplied citations are sometimes inaccurate, and current evaluations score citations at generation time or against text passages: no existing benchmark evaluates whether a system can verify and correct a page-level citation already attached to an answer. We propose AtomCite, an agentic framework that parses an answer into claims, checks each claim against the image of its cited page, and applies a deterministic repair policy. To evaluate it, we introduce DocCite, to our knowledge the first benchmark for systems that verify and correct page-level citations in document images. Built on MP-DocVQA and DUDE, it combines 928 validated injected instances with 2,468 candidate natural errors harvested from frontier- and efficiency-tier models, of which a two-annotator audit confirms 1,909 as genuine errors. Primary labels are assigned deterministically, not by LLM judges, with the human audit as a separate validation layer. Across three model families (Gemini, Claude, and GPT), AtomCite reaches around 93% binary verification accuracy on the injected benchmark, significantly outperforming every OCR-only condition, including a compute-matched control, and exceeding every prior text-based baseline given the same OCR text. Its repair policy lifts citation precision on the injected mix from a constructed 34% to 87-90% while retaining over 90% of correct claims. AtomCite also transfers: with frozen prompts and zero training, it raises the hallucination-detection scores of two open 7-8B models on five public benchmarks above the same models prompted as direct judges. Finally, the audit shows that noise in automatic labels biases measured verifier accuracy and can reverse system rankings, so evaluations relying only on synthetic or automatic labels risk mismeasuring verification capability.

0.24

Understanding before verifying: Claim normalization for automated citation verification

cs.IR ·2026-08-31 ·Yifan He, Mengjia Wu, Siming Deng, Yi Zhang
similarity 0.57
Abstract

Citation accuracy has been studied for decades because of its importance to research reliability. Content-level citation verification assesses the reliability of scholarly claims. Recent work adopts a two-stage retrieval-classification framework inherited from fact-checking. However, this design overlooks the complexity of the raw citing claim and introduces three issues into the verification system, namely scope mismatch, perspective mismatch, and proposition entanglement. These issues increase the difficulty of retrieval and classification, thereby limiting model performance. Motivated by this gap, we propose claim normalization, which applies three rewriting strategies to the raw citing claim before retrieval and classification, allowing each downstream model to perform a single, well-defined task. Building on this method, we develop Claim-Normalized Citation Verification (CNCV), a new three-stage framework consisting of claim normalization, evidence retrieval with grounding, and citation classification. We evaluate CNCV across 18 classifiers using a factorial experiment on human-annotated citation instances. Compared with the prior two-stage framework, CNCV improves macro F1 by an average of 12% for encoders and 10% for generative LLMs, driven by improved evidence quality, the dominant factor identified in our experiments. Evidence retrieved from automatically normalized claims yields downstream classification performance statistically equivalent to that obtained with manually annotated evidence.

0.25

Rethinking Inline Citation Verification in Scholarly Communication

cs.HC ·2026-10-04 ·Xinrui Fang, Reese Fairchild, Nasi Wang, Anran Xu, Simo Hosio +2 more
survey similarity 0.57
Abstract

Inline citations are central to scholarly communication, yet verifying their use is becoming increasingly challenging because of growing review pressures. We conducted a mixed-methods exploratory study to investigate how reviewers verify inline citations, the factors shaping their verification practices, and how they envision GenAI supporting this process. Across interviews (n=12) and a survey (n=203) of reviewers from HCI and AI venues, we found that reviewers' perceptions of citation importance, verification practices, and desired AI autonomy varied across citation types, reviewer characteristics, and research backgrounds. These findings highlight the need for adaptive support tailored to different citation types and reviewer practices, while revealing diverse preferences regarding the use of GenAI for this process. Moreover, effective citation verification should involve collaboration among reviewers, authors, and the broader research community, rather than relying solely on individual reviewers.

0.11