StereoTales shows that all tested LLMs emit harmful stereotypes in open-ended stories, with associations adapting to prompt language and targeting locally salient groups rather than transferring uniformly across languages.
hub
URL https://aclantholo gy.org/2023.acl-long.557/
35 Pith papers cite this work. Polarity classification is still indexing.
35
Pith papers citing it
hub tools
co-cited works
years
2026 35representative citing papers
EDEN adaptively sets branching factor proportional to next-token entropy, achieving better accuracy per expansion than fixed beam search while providing a proof that monotone entropy-based branching outperforms any fixed budget allocation.
COPYCOP identifies copycat GNNs by matching their node embeddings despite architectural differences and adversarial transformations, backed by theoretical guarantees and tests on 14 datasets across 5 architectures.
MedMosaic is a large-scale medical audio QA benchmark that shows even state-of-the-art models like Gemini-2.5-pro reach only about 68% accuracy on diverse clinical audio scenarios.
ReformIR adaptively prioritizes reformulations and documents with a surrogate model guided by ranker feedback to boost recall while suppressing drift under fixed reranking budgets.
AI Overviews and Gemini retrieve substantially different sources than traditional Google search (Jaccard similarity <0.2), favor Google-owned content, appear for 51.5% of queries especially controversial ones, and are less consistent across repeated or slightly edited queries.
FlowBot automatically induces LLM workflows through bilevel optimization with textual gradients, achieving competitive performance against human-crafted baselines.
Skill-RAG detects retrieval failure states from hidden representations and routes to one of four corrective skills to raise accuracy on persistent hard cases in open-domain QA and reasoning benchmarks.
Test-time LLM feedback refines query embeddings to deliver up to 25% relative gains on zero-shot literature search, intent detection, and related benchmarks.
An extended annotation scheme with new categories and attributes plus a Gemma-300M-based multi-head classifier achieves 81.6% macro F1 on personal fact classification, outperforming few-shot LLM baselines by nearly 9 points with lower compute.
Teachers' views on AI benefits and risks vary widely across 55 countries, but LLMs compress these differences, overestimate both sides, and show little improvement from country prompting or better reasoning.
Sem-ECE is an asymptotically unbiased calibration error estimator for open-ended QA that uses semantic sampling of answers to derive confidence from class frequencies, with two variants that diverge on hard questions.
MedVIGIL introduces a clinician-supervised benchmark showing medical VLMs frequently give fluent answers on broken visual evidence, with top models 14 points below human radiologists on the composite score.
Fine-tuned GPT-4o reaches state-of-the-art on grammatical error correction while reference-based metrics underestimate performance by missing 73.76 percent of valid or superior outputs.
HyperEyes presents a parallel multimodal search agent using dual-grained efficiency-aware RL with a new TRACE reward and IMEB benchmark, claiming 9.9% higher accuracy and 5.3x fewer tool calls than prior open-source agents.
CoE applies vision-language models directly to document screenshots to deliver pixel-level bounding-box attribution for evidence in iterative retrieval-augmented generation, outperforming text baselines on visual-layout tasks.
SIEVES improves selective prediction coverage up to 3x on OOD VQA benchmarks by training a selector on visual localization quality, generalizing across datasets and proprietary reasoners without specific adaptation.
An endpoint point-relation classifier followed by decoding to interval relations achieves 70.1% temporal awareness on TempEval-3, setting a new state-of-the-art for the full set of fine-grained temporal relations.
SimpleTES scales test-time evaluation in LLMs to discover state-of-the-art solutions on 21 scientific problems across six domains, outperforming frontier models and optimization pipelines with examples like 2x faster LASSO and new Erdos constructions.
Anonymization placement in RAG—at the dataset or at the generated answer—creates observable differences in privacy protection versus response utility.
GRM ranks Mel bands by attack contribution versus utility sensitivity, perturbs a subset, and learns a universal perturbation to reach 88.46% average jailbreak success rate with improved attack-utility trade-off on four audio LLMs.
LLMLingua prompt compression yields up to 18% end-to-end LLM speedups with unchanged quality when prompt length, ratio, and hardware align, plus an open profiler to predict the break-even point.
GRC unifies generation, retrieval, and compression in LLMs via meta latent tokens for single-pass execution with modular flexibility.
Sycophancy is a boundary failure between social alignment and epistemic integrity, captured by a three-condition framework plus taxonomy of targets, mechanisms, and severity.
citing papers explorer
-
When Helpfulness Becomes Sycophancy: Sycophancy is a Boundary Failure Between Social Alignment and Epistemic Integrity in Large Language Models
Sycophancy is a boundary failure between social alignment and epistemic integrity, captured by a three-condition framework plus taxonomy of targets, mechanisms, and severity.
-
pAI/MSc: ML Theory Research with Humans on the Loop
pAI/MSc is a customizable multi-agent system that reduces human steering by orders of magnitude when turning a hypothesis into a literature-grounded, mathematically established, experimentally supported manuscript draft in ML theory.