Imagen de Portada

Interpreting Negation in GPT-2: Layer- and Head-Level Causal Analysis

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Mofael, Abdullah Al, Kuhn, Lisa M., Alkadi, Ghassan, Yang, Kuo-Pao
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

Mechanistic Interpretability of GPT-2: Lexical and Contextual Layers in Sentiment Analysis
por: Hatua, Amartya
Publicado: (2025)

FourierKAN outperforms MLP on Text Classification Head Fine-tuning
por: Imran, Abdullah Al, et al.
Publicado: (2024)

Positional Cognitive Specialization: Where Do LLMs Learn To Comprehend and Speak Your Language?
por: Salim, Luis Frentzen, et al.
Publicado: (2026)

ChatGPT Evaluation on Sentence Level Relations: A Focus on Temporal, Causal, and Discourse Relations
por: Chan, Chunkit, et al.
Publicado: (2023)

The Mechanics of Conceptual Interpretation in GPT Models: Interpretative Insights
por: Aljaafari, Nura, et al.
Publicado: (2024)

Is ChatGPT the Future of Causal Text Mining? A Comprehensive Evaluation and Analysis
por: Takayanagi, Takehiro, et al.
Publicado: (2024)

Large Language Model Pruning
por: Huang, Hanjuan, et al.
Publicado: (2024)

Interpreting Transformers Through Attention Head Intervention
por: Kadem, Mason, et al.
Publicado: (2026)

Causal Interventions on Causal Paths: Mapping GPT-2's Reasoning From Syntax to Semantics
por: Lee, Isabelle, et al.
Publicado: (2024)

Generating Hard-Negative Out-of-Scope Data with ChatGPT for Intent Classification
por: Li, Zhijian, et al.
Publicado: (2024)

Reversed Attention: On The Gradient Descent Of Attention Layers In GPT
por: Katz, Shahar, et al.
Publicado: (2024)

CausalDetox: Causal Head Selection and Intervention for Language Model Detoxification
por: Wang, Yian, et al.
Publicado: (2026)

Single and Multi-Hop Question-Answering Datasets for Reticular Chemistry with GPT-4-Turbo
por: Rampal, Nakul, et al.
Publicado: (2024)

Interpretability at Scale: Identifying Causal Mechanisms in Alpaca
por: Wu, Zhengxuan, et al.
Publicado: (2023)

SpeLLM: Character-Level Multi-Head Decoding
por: Ben-Artzy, Amit, et al.
Publicado: (2025)

You can remove GPT2's LayerNorm by fine-tuning
por: Heimersheim, Stefan
Publicado: (2024)

BengaliFig: A Low-Resource Challenge for Figurative and Culturally Grounded Reasoning in Bengali
por: Sefat, Abdullah Al
Publicado: (2025)

Feature Extraction and Analysis for GPT-Generated Text
por: Selvioğlu, A., et al.
Publicado: (2025)

Freeze Deep, Train Shallow: Interpretable Layer Allocation for Continued Pre-Training
por: Wu, Yu-Hang, et al.
Publicado: (2026)

Debiasing CLIP: Interpreting and Correcting Bias in Attention Heads
por: Yeo, Wei Jie, et al.
Publicado: (2025)

Evaluating the Predictive Capacity of ChatGPT for Academic Peer Review Outcomes Across Multiple Platforms
por: Thelwall, Mike, et al.
Publicado: (2024)

A comparison of Human, GPT-3.5, and GPT-4 Performance in a University-Level Coding Course
por: Yeadon, Will, et al.
Publicado: (2024)

ExaGPT: Example-Based Machine-Generated Text Detection for Human Interpretability
por: Koike, Ryuto, et al.
Publicado: (2025)

From Interpretability to Performance: Optimizing Retrieval Heads for Long-Context Language Models
por: Ma, Youmi, et al.
Publicado: (2026)

Preference Heads in Large Language Models: A Mechanistic Framework for Interpretable Personalization
por: Zhang, Weixu, et al.
Publicado: (2026)

ReBeCA: Unveiling Interpretable Behavior Hierarchy behind the Iterative Self-Reflection of Language Models with Causal Analysis
por: Yan, Tianqiang, et al.
Publicado: (2026)

The Qiyas Benchmark: Measuring ChatGPT Mathematical and Language Understanding in Arabic
por: Al-Khalifa, Shahad, et al.
Publicado: (2024)

Interpretable Next-token Prediction via the Generalized Induction Head
por: Kim, Eunji, et al.
Publicado: (2024)

Thunder-NUBench: A Benchmark for LLMs' Sentence-Level Negation Understanding
por: So, Yeonkyoung, et al.
Publicado: (2025)

TermGPT: Multi-Level Contrastive Fine-Tuning for Terminology Adaptation in Legal and Financial Domain
por: Sun, Yidan, et al.
Publicado: (2025)

InfraGPT Smart Infrastructure: An End-to-End VLM-Based Framework for Detecting and Managing Urban Defects
por: Mohamed, Ibrahim Sheikh, et al.
Publicado: (2025)

SkipGPT: Dynamic Layer Pruning Reinvented with Token Awareness and Module Decoupling
por: Zhao, Anhao, et al.
Publicado: (2025)

Geometric Interpretation of Layer Normalization and a Comparative Analysis with RMSNorm
por: Gupta, Akshat, et al.
Publicado: (2024)

Flash Interpretability: Decoding Specialised Feature Neurons in Large Language Models with the LM-Head
por: Davies, Harry J
Publicado: (2025)

EmoBang: Detecting Emotion From Bengali Texts
por: Maruf, Abdullah Al, et al.
Publicado: (2025)

Evaluating Subword Tokenization Techniques for Bengali: A Benchmark Study with BengaliBPE
por: Patwary, Firoj Ahmmed, et al.
Publicado: (2025)

ShortGPT: Layers in Large Language Models are More Redundant Than You Expect
por: Men, Xin, et al.
Publicado: (2024)

Mechanism and Emergence of Stacked Attention Heads in Multi-Layer Transformers
por: Musat, Tiberiu
Publicado: (2024)

K-ON: Stacking Knowledge On the Head Layer of Large Language Model
por: Guo, Lingbing, et al.
Publicado: (2025)

PubMedCausal: A Span-Level Annotated Corpus for Causal Relation Extraction in Biomedical Text
por: Kunle-John, Ifeoluwa, et al.
Publicado: (2026)