Imagen de Portada

Memorization, Emergence, and Explaining Reversal Failures: A Controlled Study of Relational Semantics in LLMs

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Zhu, Yihua, Liu, Qianying, Wang, Jiaxin, Cheng, Fei, Liu, Chaoran, Aizawa, Akiko, Kurohashi, Sadao, Shimodaira, Hidetoshi
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Reasoning Depth and Environment Complexity: A Controlled Study of RLVR Data Allocation across Logical Reasoning Tasks
por: Zhu, Yihua, et al.
Publicado: (2026)

Beyond Chains: Bridging Large Language Models and Knowledge Bases in Complex Question Answering
por: Zhu, Yihua, et al.
Publicado: (2025)

Block-Diagonal Orthogonal Relation and Matrix Entity for Knowledge Graph Embedding
por: Zhu, Yihua, et al.
Publicado: (2024)

Assessing Agentic Large Language Models in Multilingual National Bias
por: Liu, Qianying, et al.
Publicado: (2025)

3D Rotation and Translation for Hyperbolic Knowledge Graph Embedding
por: Zhu, Yihua, et al.
Publicado: (2023)

Language Lives in Sparse Dimensions: Toward Interpretable and Efficient Multilingual Control for Large Language Models
por: Zhong, Chengzhi, et al.
Publicado: (2025)

Mechanistic Diagnostics of Spatial Lexical Bias in Multimodal Large Language Model Spatial Reasoning
por: Ma, Chuang, et al.
Publicado: (2026)

A Joint Neural Baseline for Concept, Assertion, and Relation Extraction from Clinical Text
por: Cheng, Fei, et al.
Publicado: (2026)

Beyond English-Centric LLMs: What Language Do Multilingual Language Models Think in?
por: Zhong, Chengzhi, et al.
Publicado: (2024)

BIS Reasoning 1.0: The First Large-Scale Japanese Benchmark for Belief-Inconsistent Syllogistic Reasoning
por: Nguyen, Ha-Thanh, et al.
Publicado: (2025)

Reformulating Domain Adaptation of Large Language Models as Adapt-Retrieve-Revise: A Case Study on Chinese Legal Domain
por: wan, Zhen, et al.
Publicado: (2023)

EMS: Efficient and Effective Massively Multilingual Sentence Embedding Learning
por: Mao, Zhuoyuan, et al.
Publicado: (2022)

Understanding Higher-Order Correlations Among Semantic Components in Embeddings
por: Oyama, Momose, et al.
Publicado: (2024)

Knowledge Sanitization of Large Language Models
por: Ishibashi, Yoichi, et al.
Publicado: (2023)

Norm of Mean Contextualized Embeddings Determines their Variance
por: Yamagiwa, Hiroaki, et al.
Publicado: (2024)

Causal Tree Extraction from Medical Case Reports: A Novel Task for Experts-like Text Comprehension
por: Yahata, Sakiko, et al.
Publicado: (2025)

Rapidly Developing High-quality Instruction Data and Evaluation Benchmark for Large Language Models with Minimal Human Effort: A Case Study on Japanese
por: Sun, Yikun, et al.
Publicado: (2024)

AcTED: Automatic Acquisition of Typical Event Duration for Semi-supervised Temporal Commonsense QA
por: Virgo, Felix, et al.
Publicado: (2024)

RecMind: Japanese Movie Recommendation Dialogue with Seeker's Internal State
por: Kodama, Takashi, et al.
Publicado: (2024)

Quantifying Lexical Semantic Shift via Unbalanced Optimal Transport
por: Kishino, Ryo, et al.
Publicado: (2024)

Shimo Lab at "Discharge Me!": Discharge Summarization by Prompt-Driven Concatenation of Electronic Health Record Sections
por: He, Yunzhen, et al.
Publicado: (2024)

Revisiting Cosine Similarity via Normalized ICA-transformed Embeddings
por: Yamagiwa, Hiroaki, et al.
Publicado: (2024)

Measuring Affinity between Attention-Head Weight Subspaces via the Projection Kernel
por: Yamagiwa, Hiroaki, et al.
Publicado: (2026)

Language Model Maps for Prompt-Response Distributions via Log-Likelihood Vectors
por: Takase, Yusuke, et al.
Publicado: (2026)

Axis Tour: Word Tour Determines the Order of Axes in ICA-transformed Embeddings
por: Yamagiwa, Hiroaki, et al.
Publicado: (2024)

Are Emotions Arranged in a Circle? Geometric Analysis of Emotion Representations via Hyperspherical Contrastive Learning
por: Yamauchi, Yusuke, et al.
Publicado: (2026)

Unsupervised Domain Adaptation for Keyphrase Generation using Citation Contexts
por: Boudin, Florian, et al.
Publicado: (2024)

Mapping 1,000+ Language Models via the Log-Likelihood Vector
por: Oyama, Momose, et al.
Publicado: (2025)

Likelihood Variance as Text Importance for Resampling Texts to Map Language Models
por: Oyama, Momose, et al.
Publicado: (2025)

An Analysis of Datasets, Metrics and Models in Keyphrase Generation
por: Boudin, Florian, et al.
Publicado: (2025)

Preface to the Special Issue of the TAL Journal on Scholarly Document Processing
por: Boudin, Florian, et al.
Publicado: (2025)

Predicting drug-gene relations via analogy tasks with word embeddings
por: Yamagiwa, Hiroaki, et al.
Publicado: (2024)

Encoded but Not Routed: Explaining the Table-Chart Gap in Scientific Claim Verification
por: Kumar, Sunisth, et al.
Publicado: (2026)

Harnessing PDF Data for Improving Japanese Large Multimodal Models
por: Baek, Jeonghun, et al.
Publicado: (2025)

Table-Text Alignment: Explaining Claim Verification Against Tables in Scientific Papers
por: Ho, Xanh, et al.
Publicado: (2025)

Zipfian Whitening
por: Yokoi, Sho, et al.
Publicado: (2024)

DeLTa: A Decoding Strategy based on Logit Trajectory Prediction Improves Factuality and Reasoning Ability
por: He, Yunzhen, et al.
Publicado: (2025)

Reading Comprehension using Entity-based Memory Network
por: Wang, Xun, et al.
Publicado: (2016)

Tracing the Arrow of Time: Diagnosing Temporal Information Flow in Video-LLMs
por: Han, Peitao, et al.
Publicado: (2026)

MELD-ST: An Emotion-aware Speech Translation Dataset
por: Chen, Sirou, et al.
Publicado: (2024)