:: Library Catalog

Imagen de Portada

Guardado en:

Detalles Bibliográficos
Autor principal:	Abram, Marcin
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language
Acceso en línea:	https://arxiv.org/abs/2604.04043
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Emergent Structured Representations Support Flexible In-Context Inference in Large Language Models
por: Xu, Ningyu, et al.
Publicado: (2026)

Detecting Data Contamination in LLMs via In-Context Learning
por: Zawalski, Michał, et al.
Publicado: (2025)

Epistemic Blinding: An Inference-Time Protocol for Auditing Prior Contamination in LLM-Assisted Analysis
por: Cuccarese, Michael
Publicado: (2026)

Toward Evaluation Frameworks for Multi-Agent Scientific AI Systems
por: Abram, Marcin
Publicado: (2026)

Context Steering: Controllable Personalization at Inference Time
por: He, Jerry Zhi-Yang, et al.
Publicado: (2024)

ChunkKV: Semantic-Preserving KV Cache Compression for Efficient Long-Context LLM Inference
por: Liu, Xiang, et al.
Publicado: (2025)

Are Emergent Abilities in Large Language Models just In-Context Learning?
por: Lu, Sheng, et al.
Publicado: (2023)

Why does in-context learning fail sometimes? Evaluating in-context learning on open and closed questions
por: Li, Xiang, et al.
Publicado: (2024)

Semantic Reranking at Inference Time for Hard Examples in Rhetorical Role Labeling
por: Belfathi, Anas, et al.
Publicado: (2026)

Automated Semantic Rules Detection (ASRD) for Emergent Communication Interpretation
por: Vanderplaetse, Bastien, et al.
Publicado: (2026)

Emergent Misalignment via In-Context Learning: Narrow in-context examples can produce broadly misaligned LLMs
por: Afonin, Nikita, et al.
Publicado: (2025)

Laying Anchors: Semantically Priming Numerals in Language Modeling
por: Sharma, Mandar, et al.
Publicado: (2024)

Cross-Context Verification: Hierarchical Detection of Benchmark Contamination through Session-Isolated Analysis
por: Song, Tae-Eun
Publicado: (2026)

Towards Human Cognition: Visual Context Guides Syntactic Priming in Fusion-Encoded Models
por: Xiao, Bushi, et al.
Publicado: (2025)

Emergent Visual-Semantic Hierarchies in Image-Text Representations
por: Alper, Morris, et al.
Publicado: (2024)

Emergent Representations of Program Semantics in Language Models Trained on Programs
por: Jin, Charles, et al.
Publicado: (2023)

Experimental Contexts Can Facilitate Robust Semantic Property Inference in Language Models, but Inconsistently
por: Misra, Kanishka, et al.
Publicado: (2024)

Semantic Containment as a Fundamental Property of Emergent Misalignment
por: Saxena, Rohan
Publicado: (2026)

Accelerating Inference of Retrieval-Augmented Generation via Sparse Context Selection
por: Zhu, Yun, et al.
Publicado: (2024)

Efficient Long-Context LLM Inference via KV Cache Clustering
por: Hu, Jie, et al.
Publicado: (2025)

CAP: Data Contamination Detection via Consistency Amplification
por: Zhao, Yi, et al.
Publicado: (2024)

Steering Vector Fields for Context-Aware Inference-Time Control in Large Language Models
por: Li, Jiaqian, et al.
Publicado: (2026)

Emergent Semantic Role Understanding in Language Models
por: Griffiths, Carla, et al.
Publicado: (2026)

Autoencoding-Free Context Compression for LLMs via Contextual Semantic Anchors
por: Liu, Xin, et al.
Publicado: (2025)

Focus-dLLM: Accelerating Long-Context Diffusion LLM Inference via Confidence-Guided Context Focusing
por: Long, Lingkun, et al.
Publicado: (2026)

GMSA: Enhancing Context Compression via Group Merging and Layer Semantic Alignment
por: Tang, Jiwei, et al.
Publicado: (2025)

DynSplit-KV: Dynamic Semantic Splitting for KVCache Compression in Efficient Long-Context LLM Inference
por: Ye, Jiancai, et al.
Publicado: (2026)

Is In-Context Learning a Type of Error-Driven Learning? Evidence from the Inverse Frequency Effect in Structural Priming
por: Zhou, Zhenghao, et al.
Publicado: (2024)

Dynamic Memory Compression: Retrofitting LLMs for Accelerated Inference
por: Nawrot, Piotr, et al.
Publicado: (2024)

SlimInfer: Accelerating Long-Context LLM Inference via Dynamic Token Pruning
por: Long, Lingkun, et al.
Publicado: (2025)

LogitTrace: Detecting Benchmark Contamination via Layerwise Logit Trajectories
por: He, Zirui, et al.
Publicado: (2025)

Interpretable Semantic Gradients in SSD: A PCA Sweep Approach and a Case Study on AI Discourse
por: Plisiecki, Hubert, et al.
Publicado: (2026)

Exploiting Sparsity for Long Context Inference: Million Token Contexts on Commodity GPUs
por: Synk, Ryan, et al.
Publicado: (2025)

Exploring In-Context Learning for Frame-Semantic Parsing
por: Garat, Diego, et al.
Publicado: (2025)

Speakers Fill Lexical Semantic Gaps with Context
por: Pimentel, Tiago, et al.
Publicado: (2020)

Superlatives in Context: Modeling the Implicit Semantics of Superlatives
por: Pyatkin, Valentina, et al.
Publicado: (2024)

DIESEL -- Dynamic Inference-Guidance via Evasion of Semantic Embeddings in LLMs
por: Ganon, Ben, et al.
Publicado: (2024)

Towards Contamination Resistant Benchmarks
por: Musawi, Rahmatullah, et al.
Publicado: (2025)

Contamination Report for Multilingual Benchmarks
por: Ahuja, Sanchit, et al.
Publicado: (2024)

Divergence Decoding: Inference-Time Unlearning via Auxiliary Models
por: Merchant, Humzah, et al.
Publicado: (2026)