HFEPX Daily Papers for 2026-05-21

Researcher Quick Triage

Use this archive page for time-slice monitoring (what changed in evaluation methods, metrics, and protocol quality this period). Quality band: High .

• 2 papers report explicit quality controls for this archive period.
• Prioritize papers with both benchmark and metric anchors for reliable longitudinal comparisons.

Primary action: Use this slice for trend comparison: review top papers first, then validate shifts in the protocol matrix.

Why This Time Slice Matters

• Use this archive slice to monitor protocol drift and shifts in evaluation methods over 2026-05-21.

Protocol Takeaways For This Period

• Evaluation modes for this slice cluster around automatic_metrics, simulation_env.

Start Here (Highest-Signal Papers In This Slice)

Ranked by protocol completeness and evidence density for faster period-over-period review.

• Vector Policy Optimization: Training for Diversity Improves Test-Time Search

  May 21, 2026 · Citations: 0 · Score: 6.5

  Eval: Automatic Metrics · Metrics: Pass@k

• Evaluating Commercial AI Chatbots as News Intermediaries

  May 21, 2026 · Citations: 0 · Score: 6.5

  Eval: Automatic Metrics · Metrics: Accuracy

• More Context, Larger Models, or Moral Knowledge? A Systematic Study of Schwartz Value Detection in Political Texts

  May 21, 2026 · Citations: 0 · Score: 6.5

  Eval: Automatic Metrics · Metrics: F1, F1 macro

• DeferMem: Query-Time Evidence Distillation via Reinforcement Learning for Long-Term Memory QA

  May 21, 2026 · Citations: 0 · Score: 6.5

  Eval: Automatic Metrics · Metrics: Accuracy, Recall

• Ratchet: A Minimal Hygiene Recipe for Self-Evolving LLM Agents

  May 21, 2026 · Citations: 0 · Score: 6.5

  Eval: Automatic Metrics · Metrics: Pass@1

• Psy-Chronicle:A Structured Pipeline for Synthesizing Long-Horizon Campus Psychological Counseling Dialogues

  May 21, 2026 · Citations: 0 · Score: 6.5

  Eval: Simulation Env · Metrics: Recall

Protocol Matrix (Top 10)

Quickly compare method ingredients across this archive slice.

Papers In This Archive Slice

• When AI Takes Sides on Questions of Faith: Persistent Asymmetries in AI-Mediated Faith Guidance

  Brett Israelsen, Sheryl Carty, Josh Coates, Nancy Fulda, Julie Park · May 21, 2026 · Citations: 0

  Pairwise Preference

  We tested 20 commercial and open-source language models across 182 religion pairings using a human-verified LLM-as-judge framework.

• Reducing Political Manipulation with Consistency Training

  Long Phan, Devin Kim, Alexander Pan, Alice Blair, Adam Khoja · May 21, 2026 · Citations: 0

  We show that PCT preserves overall helpfulness, substantially reduces covert political bias, and generalizes to held-out benchmarks.

• Chinese sensorimotor and embodiment norms for 3,000 lexicalized concepts

  Jing Chen, Gábor Parti, Yin Zhong, Chu-Ren Huang, Marco Marelli · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Tokenisation via Convex Relaxations

  Jan Tempus, Philip Whittington, Craig W. Schmidt, Dennis Komm, Tiago Pimentel · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Vector Policy Optimization: Training for Diversity Improves Test-Time Search

  Ryan Bahlous-Boldi, Isha Puri, Idan Shenfeld, Akarsh Kumar, Mehul Damani · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Evaluating Commercial AI Chatbots as News Intermediaries

  Mirac Suzgun, Emily Shen, Federico Bianchi, Alexander Spangher, Thomas Icard · May 21, 2026 · Citations: 0

  We present a 14-day (February 9-22, 2026) evaluation of six AI chatbots (Gemini 3 Flash and Pro, Grok 4, Claude 4.5 Sonnet, GPT-5 and GPT-4o mini) on 2,100 factual questions derived from same-day BBC News reporting across six regional…

• Understanding Data Temporality Impact on Large Language Models Pre-training

  Pilchen Hippolyte, Fabre Romain, Signe Talla Franck, Perez Patrick, Grave Edouard · May 21, 2026 · Citations: 0

  First, we introduce a comprehensive benchmark of over 7,000 temporally grounded questions and an evaluation protocol that enables analysis of whether models correctly associate facts with their corresponding time periods.

• ChronoMedKG: A Temporally-Grounded Biomedical Knowledge Graph and Benchmark for Clinical Reasoning

  Md Shamim Ahmed, Farzaneh Firoozbakht, Lukas Galke Poech, Jan Baumbach, Richard Röttger · May 21, 2026 · Citations: 0

  Multi Agent

  The graph is constructed through a disease-autonomous multi-agent pipeline in which multiple frontier LLMs independently extract knowledge from PubMed and PMC literature.

• Beyond Acoustic Emotion Recognition: Multimodal Pathos Analysis in Political Speech Using LLM-Based and Acoustic Emotion Models

  Juergen Dietrich · May 21, 2026 · Citations: 0

  Multi Agent

  We investigate whether acoustic emotion recognition models can serve as proxies for the Pathos dimension in political speech analysis, as operationalised by the TRUST multi-agent large language model (LLM) pipeline.

• AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild

  Baiyu Chen, Zechen Li, Wilson Wongso, Lihuan Li, Xiachong Lin · May 21, 2026 · Citations: 0

  As wearable and mobile devices become increasingly embedded in daily life, they offer a practical way to continuously sense human motion in the wild.

• AMEL: Accumulated Message Effects on LLM Judgments

  Sid-ali Temkit · May 21, 2026 · Citations: 0

  Tool Use

  Across 75,898 API calls to 11 models from 4 providers (OpenAI, Anthropic, Google, and four open-source models), we present identical test items in isolation or following histories saturated with predominantly positive or negative…

• Tokenization with Split Trees

  Craig W. Schmidt, Michael Krumdick, Adam Wiemerslage, Seth Ebner, Varshini Reddy · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Self-Policy Distillation via Capability-Selective Subspace Projection

  Guangya Hao, Yitong Shang, Yunbo Long, Zhuokai Zhao, Hanxue Liang · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Moral Semantics Survive Machine Translation: Cross-Lingual Evidence from Moral Foundations Corpora

  Maciej Skorski · May 21, 2026 · Citations: 0

  Using \sim50k morally-annotated social media posts from a diverse range of topics, we apply a principled four-method validation pipeline: LaBSE cross-lingual embedding similarity, Centered Kernel Alignment (CKA), LLM-as-judge evaluation,…

• Seeing the Poem: Image-Semantic Detection of AI-Generated Modern Chinese Poetry with MLLMs

  Shanshan Wang, Fengying Ye, Hanjia Lyu, Caiwen Gou, Junchao Wu · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Whose Voice Counts? Mapping Stakeholder Perspectives on AI Through Public Submissions to the U.S. Government

  Alina Karakanta, Alex Christiansen, Tomás Dodds, Bissie Anderson, Matteo Fuoli · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Boiling the Frog: A Multi-Turn Benchmark for Agentic Safety

  Piercosma Bisconti, Matteo Prandi, Federico Pierucci, Federico Sartore, Enrico Panai · May 21, 2026 · Citations: 0

  Traditional safety benchmarks for language models evaluate generated text: whether a model outputs toxic language, reproduces bias, or follows harmful instructions.

• More Context, Larger Models, or Moral Knowledge? A Systematic Study of Schwartz Value Detection in Political Texts

  Víctor Yeste, Paolo Rosso · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• The Double Dilemma in Multi-Task Radiology Report Generation: A Gradient Dynamics Analysis and Solution

  Erjian Zhang, Yatong Hao, Liejun Wang, Zhiqing Guo · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Two is better than one: A Collapse-free Multi-Reward RLIF Training Framework

  Shourov Joarder, Diganta Sikdar, Ahsan Habib Akash, Binod Bhattarai, Prashnna Gyawali · May 21, 2026 · Citations: 0

  Long Horizon

  Reinforcement learning with verifiable rewards (RLVR) has substantially improved the reasoning ability of LLMs, but often depends on external supervision from human annotations or gold-standard solutions.

• Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents

  Asaf Yehudai, Lilach Eden, Michal Shmueli-Scheuer · May 21, 2026 · Citations: 0

  To address this gap, we present Agentic CLEAR, an automatic, dynamic, and easy-to-use evaluation framework.

• A Tutorial on Diffusion Theory: From Differential Equations to Diffusion Models

  Jiayi Fu, Yuxia Wang · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Beyond Temperature: Hyperfitting as a Late-Stage Geometric Expansion

  Meimingwei Li, Yuanhao Ding, Esteban Garces Arias, Christian Heumann · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• LANG: Reinforcement Learning for Multilingual Reasoning with Language-Adaptive Hint Guidance

  Yuchun Fan, Bei Li, Peiguang Li, Yilin Wang, Yongyu Mu · May 21, 2026 · Citations: 0

  Long Horizon

  Empirical results on challenging multilingual mathematical benchmarks reveal that LANG substantially enhances reasoning performance without compromising language consistency.

• SynAE: A Framework for Measuring the Quality of Synthetic Data for Tool-Calling Agent Evaluations

  Shuaiqi Wang, Aadyaa Maddi, Zinan Lin, Giulia Fanti · May 21, 2026 · Citations: 0

  We introduce SynAE, an evaluation framework for assessing how well synthetic benchmarks for multi-turn, tool-calling agents replicate and augment the characteristics of real data trajectories.

• One prompt is not enough: Instruction Sensitivity Undermines Embedding Model Evaluation

  Yevhen Kostiuk, Kenneth Enevoldsen · May 21, 2026 · Citations: 0

  The single-point evaluation ignores a main problem of the instruction-based approach namely: sensitivity to the phrasing of the instruction.

• Scene Abstraction for Lexical Semantics: Structured Representations of Situated Meaning

  Yejin Cho, Katrin Erk · May 21, 2026 · Citations: 0

  Pairwise Preference

  Our contributions are three-fold: (1) a structured representation framework for situated lexical meaning; (2) COCA-Scenes, a dataset of 520 usage instances across 26 keywords for distinct scene identification; and (3) empirical evidence…

• SpaceDG: Benchmarking Spatial Intelligence under Visual Degradation

  Xiaolong Zhou, Yifei Liu, Ziyang Gong, Jiarui Li, Qiyue Zhao · May 21, 2026 · Citations: 0

  Multimodal Large Language Models (MLLMs) have made rapid progress in spatial intelligence, yet existing spatial reasoning benchmarks largely assume pristine visual inputs and overlook the degradations that commonly occur in real-world…

• Search-E1: Self-Distillation Drives Self-Evolution in Search-Augmented Reasoning

  Zihan Liang, Yufei Ma, Ben Chen, Zhipeng Qian, Xuxin Zhang · May 21, 2026 · Citations: 0

  Long Horizon

  Post-training has become the dominant recipe for turning a language model into a competent search-augmented reasoning agent.

• Reflecti-Mate: A Conversational Agent for Adaptive Decision-Making Support Through System 1 and System 2 Thinking

  Morita Tarvirdians, Senthil Chandrasegaran, Hayley Hung, Catholijn M. Jonker, Catharine Oertel · May 21, 2026 · Citations: 0

  In this study, we investigate an agent designed to encourage integration by adapting to the individual user's thought patterns.

• BeLink: Biomedical Entity Linking Meets Generative Re-Ranking

  Darya Shlyk, Stefano Montanelli, Lawrence Hunter · May 21, 2026 · Citations: 0

  Our method demonstrates strong performance on multiple BEL benchmarks, yielding significant improvements in linking accuracy (3%-24%) while reducing inference time compared to the state-of-the-art.

• Polite on the Surface, Wrong in Practice: A Curated Dataset for Fixing Honorific Failures in Multilingual Bangla Generation

  Md. Asaduzzaman Shuvo, Mahedi Hasan, Md. Tashin Parvez, Azizul Haque Noman, Md. Shafayet Hossain Ovi · May 21, 2026 · Citations: 0

  To address this limitation, we introduce a novel, culturally aligned instruction-tuning dataset for BangLa Application and DialoguE generation - BLADE and benchmarking framework comprising 4,196 meticulously curated interaction pairs.

• Structured-Sparse Attention for Entity Tracking with Subquadratic Sequence Complexity

  Hangyue Zhao, Paul Caillon, Erwan Fagnou, Alexandre Allauzen · May 21, 2026 · Citations: 0

  Recent task-specific attention operators can compress deep Transformer stacks into a few layers by performing multi-hop state propagation within a single layer, but their dense evaluation remains expensive.

• In Silico Modeling of the RAMPHO Buffer: Dissociating Informational and Energetic Masking via Phonetic Entropy in Deep Neural Networks

  Stefan Bleeck · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• From Correlation to Cause: A Five-Stage Methodology for Feature Analysis in Transformer Language Models

  Caleb Munigety · May 21, 2026 · Citations: 0

  Two NLA-inspired evaluations strengthen this picture: the fifteen selective features explain only 31% of activation variance versus the SAE's 99.7%, and selectivity ratio anticorrelates with causal force (r = -0.56).

• Cohesion-6K: An Arabic Dataset for Analyzing Social Cohesion and Conflict in Online Discourse

  Aisha Ali Al-Athba, Wajdi Zaghouani · May 21, 2026 · Citations: 0

  The annotation process combines expert human judgment with model-assisted pre-labeling verified by trained annotators, achieving substantial inter-annotator agreement (Cohens kappa = 0.85).

• Assisted Counterspeech Writing at the Crossroads of Hate Speech and Misinformation

  Genoveffa Martone, Helena Bonaldi, Marco Guerini · May 21, 2026 · Citations: 0

  23 experts revise the generated CS, which are assessed via human and automatic metrics.

• DeferMem: Query-Time Evidence Distillation via Reinforcement Learning for Long-Term Memory QA

  Jianing Yin, Tan Tang · May 21, 2026 · Citations: 0

  Large language model (LLM) agents still struggle with long-term memory question answering, where answer-supporting evidence is often scattered across long conversational histories and buried in substantial irrelevant content.

• Epicure: Navigating the Emergent Geometry of Food Ingredient Embeddings

  Jakub Radzikowski, Josef Chen · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Unified Data Selection for LLM Reasoning

  Xiaoyuan Li, Yubo Ma, Chengpeng Li, Fengbin Zhu, Yiyao Yu · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Multi-Stage Training for Abusive Comment Detection in Indic Languages

  Pranshu Rastogi, Madhav Mathur, Ramaneswaran S, Kshitij Mohan · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Boundary-targeted Membership Inference Attacks on Safety Classifiers

  Anthony Hughes, Alexander Goldberg, Prince Jha, Adam Perer, Nikolaos Aletras · May 21, 2026 · Citations: 0

  Safety classifiers are essential safeguards within generative AI systems, filtering harmful content or identifying at-risk users when interacting with large language models.

• Modeling Pathology-Like Behavioral Patterns in Language Models Through Behavioral Fine-Tuning

  Nicola Milano, Davide Marocco · May 21, 2026 · Citations: 0

  Large language models are increasingly used as computational tools for modeling human-like behavior.

• TransitLM: A Large-Scale Dataset and Benchmark for Map-Free Transit Route Generation

  Hanyu Guo, Jiedong Yang, Chao Chen, Longfei Xu, Kaikui Liu · May 21, 2026 · Citations: 0

  We present TransitLM, a large-scale dataset of over 13 million transit route planning records from four Chinese cities covering 120,845 stations and 13,666 lines, released as a continual pre-training corpus and benchmark data for three…

• Pattern-and-root inflectional morphology: the Arabic broken plural

  Alexis Amid Neme, Eric Laporte · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Harder to Defend: Towards Chinese Toxicity Attacks via Implicit Enhancement and Obfuscation Rewriting

  Jingyi Kang, Junyu Lu, Bo Xu, Hongbo Wang, Linlin zong · May 21, 2026 · Citations: 0

  Red Team

  We introduce Chinese Implicit Toxicity Attack (CITA), a controlled red-team evaluation and defense-data generation framework, not a deployable evasion tool.

• IdioLink: Retrieving Meaning Beyond Words Across Idiomatic and Literal Expressions

  Kai Golan Hashiloni, Daniel Fadlon, Lior Livyatan, Ofri Hefetz, Jiahuan Pei · May 21, 2026 · Citations: 0

  We introduce IdioLink, a retrieval benchmark designed to test whether models can link idiomatic expressions to conceptually equivalent meanings expressed in literal or paraphrased forms.

• GHI: Graphormer over Conditioned Hypergraph Incidence for Aspect-Based Sentiment Analysis

  Yu Du, Wenlong Zhu, Xingze Li, Chenglong Cao, Jing Wang · May 21, 2026 · Citations: 0

  Extensive experiments on six standard ABSA benchmarks show that GHI outperforms all baselines on the SemEval domains, and multi-seed evaluations show stable improvements over strong DeBERTa.

• Survive or Collapse: The Asymmetric Roles of Data Gating and Reward Grounding in Self-Play RL

  Sophia Xiao Pu, Zhaotian Weng, Chengzhi Liu, Jayanth Srinivasa, Gaowen Liu · May 21, 2026 · Citations: 0

  Self-play reinforcement learning trains language models on their own generated tasks, co-evolving a proposer and solver without human labels.

• Audience Engagement with Arabic Women's Social Empowerment and Wellbeing: A Decadal Corpus

  Wajdi Zaghouani, Mabrouka Bessghaier, MD. Rafiul Biswas, Shimaa Amer Ibrahim · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Evaluation of Chunking Strategies for Effective Text Embedding in Low-Resource Language on Agricultural Documents

  Sovandara Chhoun, Pichdara Po, Sereiwathna Ros, Wan-Sup Cho, Saksonita Khoeurn · May 21, 2026 · Citations: 0

  For evaluation, we perform 5-fold cross-validation over 18 question-answer pairs.

• Structure Retention in Embedding Spaces as a Predictor of Benchmark Performance

  Amanda Myntti, Jenna Kanerva, Veronika Laippala, Filip Ginter · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles

  Jinyang Wu, Guocheng Zhai, Ruihan Jin, Yuhao Shen, Zhengxi Lu · May 21, 2026 · Citations: 0

  Expert Verification

  In this paper, we present Maestro (Multimodal Agent for Expert-Skill Targeted Reinforced Orchestration), a Reinforcement Learning (RL)-driven orchestration framework that reframes heterogeneous multimodal tasks as a sequential…

• Do Factual Recall Mechanisms Carry over from Text to Speech in Multimodal Language Models?

  Luca Modica, Filip Landin, Mehrdad Farahani, Livia Qian, Gabriel Skantze · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

• Ratchet: A Minimal Hygiene Recipe for Self-Evolving LLM Agents

  Xing Zhang, Yanwei Cui, Guanghui Wang, Ziyuan Li, Wei Qiu · May 21, 2026 · Citations: 0

  We introduce Ratchet, a single-agent loop in which a frozen LLM writes, retrieves, curates, and retires its own natural-language skills.

• Psy-Chronicle:A Structured Pipeline for Synthesizing Long-Horizon Campus Psychological Counseling Dialogues

  Chaogui Gou, Jiarui Liang · May 21, 2026 · Citations: 0

  Long Horizon

  Through interactive simulation between a student agent and a counselor agent, together with a structured memory integration mechanism, Psy-Chronicle generates long-horizon dialogues with continuity across counseling sessions.

• Efficient Agentic Reasoning Through Self-Regulated Simulative Planning

  Mingkai Deng, Jinyu Hou, Lara Sá Neves, Varad Pimpalkhute, Taylor W. Killian · May 21, 2026 · Citations: 0

  Long Horizon

  To test this, we develop SR^2AM (Self-Regulated Simulative Reasoning Agentic LLM), realizing both as distinct stages within an LLM's chain-of-thought, with the LLM as world model.

• Cross-Lingual Consensus: Aligning Multilingual Cultural Knowledge via Multilingual Self-Consistency

  Andrew Ivan Soegeng, Patrick Sutanto, Tan Sang Nguyen · May 21, 2026 · Citations: 0

  Critique Edit

  Evaluations on the BLEnD benchmark demonstrate that our approach significantly improves cultural alignment-boosting performance on English queries by an average of 5.03%-relying entirely on self-generated data.

• A Comparative Study of Language Models for Khmer Retrieval-Augmented Question Answering

  Sereiwathna Ros, Phannet Pov, Ratanaktepi Chhor, Kimleang Ly, Wan-Sup Cho · May 21, 2026 · Citations: 0

  We conduct a two-phase comparative evaluation.

• ArabDiscrim: A Decade-Long Arabic Facebook Corpus on Racism and Discrimination

  Wajdi Zaghouani, Shimaa Amer Ibrahim, Mabrouka Bessghaier, Houda Bouamor · May 21, 2026 · Citations: 0

  Abstract shows limited direct human-feedback or evaluation-protocol detail; use as adjacent methodological context.

Recent Archive Slices