Ivan Decostanzi, Michele Ronco, Sergio Consoli +8cs.AI cs.CL
Effective humanitarian response depends on the rapid synthesis of heterogeneous, high-volume information sources - a task that routinely exceeds human analytical capacity in the critical early hours of a crisis. We present a pipeline that combines structured disaster records from EM-DAT with unstructured documents from ReliefWeb and the European Media Monitor (EMM) to produce source-grounded disaster storylines and causal knowledge graphs supporting situational awareness for responders and analysts. Using Retrieval-Augmented Generation, the pipeline extracts structured storylines - tabular event profiles covering 17 fields, from severity and key drivers to child-sensitive impact indicators - and constructs causal knowledge graphs where each node and edge is enriched with citation-grounded explanatory narratives, enabling full traceability back to primary sources. We evaluate the system on three diverse crisis use cases through a human evaluation involving 9 domain expert and 9 non-expert evaluators. Results confirm high retrieval precision, strong faithfulness of extracted causal relations, and a clear expert preference for citation-grounded components over ungrounded alternatives. The pipeline is designed to scale to the full EM-DAT catalogue, with the goal of publicly releasing a narrative-enriched version of the database.
Video question answering systems built on vision-language models often produce timestamped claims with high confidence even when unsupported by the cited frame. This deceptive hallucination arises because timestamps imply grounding without ensuring correctness, increasing user trust but not accuracy. We introduce a pipeline that closes this loop. A retrieval-augmented language model drafts answers with per-claim timestamp citations, and each cited frame is independently re-examined before being shown to the user. We compare against a plain baseline and ablate three verification designs, evaluated on both Apple Silicon (MLX) and Google Colab (HF Transformers, CUDA). Directly asking the vision model whether a frame supports a claim fails completely (0% catch rate on 40 claims) due to sycophancy. Blind re-captioning plus a general LLM judge improves results but is unstable, oscillating between 0% and 100% flagged depending on prompt phrasing. Replacing that judge with a small natural language inference model yields a stable, interpretable verifier that catches 79% of fabricated claims on adversarial false-premise questions while leaving true claims untouched. We release the full pipeline, evaluation harness, and implementations for both Apple Silicon and Colab. Code is available at https://github.com/yogesh-iitj/grounded-video-qa.
Search-agent rewards mix answer quality, citation grounding, tool cost, and anti-hacking terms; a high score therefore need not imply that cited evidence was retrieved, and added penalties can cancel. We introduce HERALD, an offline audit that applies exact same-question interventions, separates candidate-visible from oracle information, and enumerates detector contracts before policy optimization. On four Qwen3-8B pools from HotpotQA, 2WikiMultiHopQA, and MuSiQue, $R_0$ rejects search deletion and fake IDs, but a label-free citation-laundering attack succeeds. A complete $2^3$ ablation identifies targeted strengthening of $L$---citing a corpus passage absent from the retrieved evidence---as the observed inclusion-minimal repair: $R[L]$ has zero empirical ASR with a 0.50% one-sided cluster upper bound. The gap persists across pool rules, a visible BM25 attacker, and four models; broader hardening remains vulnerable when the attack removes an oracle support-ID penalty. Under strict 5M-token matched training evaluated on 256 paired questions per benchmark, $R[L]$ meets the EM non-inferiority gate on HotpotQA and 2Wiki but not MuSiQue. Equal-suite citation precision and support recall improve by 2.02 and 1.46 points, unsupported citations fall by 1.69, and laundering attackability falls on 2Wiki and MuSiQue. Natural $L$ is not reduced, and the detector appears in only 18 of 58,368 training trajectories. HERALD thus separates robust scoring, sparse learning signal, and policy transfer.
This paper presents the MPR-CiteG framework, which achieved second place in the ScienceON AI Challenge by addressing two fundamental challenges in generative AI: inefficient retrieval and the absence of source verification. We propose a dual-component system, termed MPR-CiteG, in which the Multi-Portfolio Retriever (MPR) efficiently retrieves diverse and relevant information, while the Citation-Grounded Generation (CiteG) module ensures that every generated output remains factually consistent and explicitly attributed to its source. MPR-CiteG represents a significant step toward building more trustworthy and accurate LLMs that are not only capable of generating information but also of grounding their responses in reliable evidence, thereby mitigating common issues like model hallucination. Extensive experiments on the challenge dataset validate the effectiveness and reliability of our approach. Our code is available at https://github.com/2noweyh/MPR-citeG.
Large language models are increasingly used to assist scientific reading, but existing evaluation methods often fail to detect whether answers are supported by verifiable citations. We introduce ResearchQA, a benchmark of 6,211 single-paper question-answer pairs from 494 open-access papers spanning eight domains and four question types: lookup, comprehension, multi-hop, and adversarial. ResearchQA is designed for citation-grounded evaluation: it permits multiple valid supporting passages for a claim and rewards grounded refusal when the source paper does not support an answer. We evaluate eight leading closed- and open-weight models in a citation-grounded chat-with-paper setting using a deterministic citation matcher and an LLM-based rubric evaluator. Citation-based metrics separate systems more clearly than LLM-evaluator scores: section coverage and citation accuracy vary substantially across models, while evaluator scores remain tightly compressed. We further find that open-weight models approach the best closed-model citation accuracy while achieving 3 to 6 times lower per-example latency. We release the benchmark, evaluation harness, and evaluator prompt.
Legal benchmarks typically score final answers even when models also state legal authority. We test whether answer correctness can serve as a proxy for authority grounding. Under ordinary reasoning prompts that did not request statutory citations, four LLMs spontaneously produced authority markers across 238 Taiwan bar-examination items. Because each item has a verified governing provision, we automatically audit answer correctness and authority grounding jointly. The two dimensions dissociate in both directions. In criminal law, 24.0--42.4\% of valid responses were answer-correct but missed the gold authority, while 15.2--21.7\% were answer-incorrect but cited it. A separate statutory-retrieval probe and a permissive citation-abstention intervention further show that answer and citation behavior can move separately at the output level. Because this mismatch arises without adversarial or inconsistency-inducing prompting, answer-only scoring treats naturally occurring gold-authority misses as complete benchmark successes. Because statutory authority is structurally extractable and externally verifiable, the failure can be measured automatically. A preliminary PRC civil-law extension also observes citation-unrequested authority marking, motivating a full cross-jurisdictional joint audit. We therefore propose joint answer--authority evaluation for statute-grounded legal benchmarks.
Large language model (LLM) agents are increasingly embedded in scientific workflows for literature analysis, drafting, and review. Existing systems advance autonomous discovery and manuscript generation, but do not resolve the governance problem that arises when ideas, methods, results, and claims propagate through AI-assisted workflows without mandatory human approval or artifact-level traceability. This paper proposes Paper Pilot, a human-in-the-loop expert system for evidence-traceable scientific manuscript generation in applied sciences. It adapts the Collaborative Agent Reasoning Engineering (CARE) methodology to manuscript development through manuscript-owner approval gates, explicit no-pass criteria, claim classification, audit logging, advisory LLM review, and evidence-locked revision control. The framework defines eight approval gates across the idea-to-claim pipeline and distinguishes literature-grounded from artifact-grounded claims, requiring reported numbers and interpretations to remain traceable to approved evidence; its system prompt is openly released for deployment in ChatGPT, Gemini, Claude, or institutional LLM environments. As a first empirical validation, we evaluate the citation-grounding layer with a controlled, mechanically scored benchmark (two commercial LLMs, real arXiv papers, no LLM judge): under coverage pressure ungated drafters fabricated up to 25% of their citations and never flagged an evidence gap, whereas the same models under Paper Pilot's evidence-locked rules produced zero fabricated citations and surfaced the planted gaps as explicit placeholders. Preliminary results for result grounding, revision, and adversarial robustness point the same way; full evaluation is left to future work. Paper Pilot positions LLM-assisted writing as a controlled human-AI decision-support process rather than a fully autonomous authorship pipeline.