Image de couverture de livre

Jump to Conclusions: Short-Cutting Transformers With Linear Transformations

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Din, Alexander Yom, Karidi, Taelin, Choshen, Leshem, Geva, Mor
Format:	Preprint
Publié:	2023
Sujets:	Computation and Language
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Assessing the Role of Lexical Semantics in Cross-lingual Transfer through Controlled Manipulations
par: Ilani, Roy, et autres
Publié: (2024)

Locally Measuring Cross-lingual Lexical Alignment: A Domain and Word Level Perspective
par: Karidi, Taelin, et autres
Publié: (2024)

A Nurse is Blue and Elephant is Rugby: Cross Domain Alignment in Large Language Models Reveal Human-like Patterns
par: Yehudai, Asaf, et autres
Publié: (2024)

The Hidden Space of Transformer Language Adapters
par: Alabi, Jesujoba O., et autres
Publié: (2024)

Detecting (Un)answerability in Large Language Models with Linear Directions
par: Lavi, Maor Juliet, et autres
Publié: (2025)

Estimating Knowledge in Large Language Models Without Generating a Single Token
par: Gottesman, Daniela, et autres
Publié: (2024)

Inferring Functionality of Attention Heads from their Parameters
par: Elhelo, Amit, et autres
Publié: (2024)

Can Gradient Descent Simulate Prompting?
par: Zhang, Eric, et autres
Publié: (2025)

The ShareLM Collection and Plugin: Contributing Human-Model Chats for the Benefit of the Community
par: Don-Yehiya, Shachar, et autres
Publié: (2024)

Naturally Occurring Feedback is Common, Extractable and Useful
par: Don-Yehiya, Shachar, et autres
Publié: (2024)

Fuse to Forget: Bias Reduction and Selective Memorization through Model Fusion
par: Zaman, Kerem, et autres
Publié: (2023)

A Hitchhiker's Guide to Scaling Law Estimation
par: Choshen, Leshem, et autres
Publié: (2024)

One Jump Is All You Need: Short-Cutting Transformers for Early Exit Prediction with One Jump to Fit All Exit Levels
par: Seshadri, Amrit Diggavi
Publié: (2025)

Instructions Shape Production of Language, not Processing
par: Waldis, Andreas, et autres
Publié: (2026)

Does Mapo Tofu Contain Coffee? Probing LLMs for Food-related Cultural Knowledge
par: Zhou, Li, et autres
Publié: (2024)

Constructing Interpretable Features from Compositional Neuron Groups
par: Shafran, Or, et autres
Publié: (2025)

Eliciting Textual Descriptions from Representations of Continuous Prompts
par: Ramati, Dana, et autres
Publié: (2024)

Narrowing the Knowledge Evaluation Gap: Open-Domain Question Answering with Multi-Granularity Answers
par: Yona, Gal, et autres
Publié: (2024)

Can Large Language Models Faithfully Express Their Intrinsic Uncertainty in Words?
par: Yona, Gal, et autres
Publié: (2024)

Hallucinations Undermine Trust; Metacognition is a Way Forward
par: Yona, Gal, et autres
Publié: (2026)

Mediocrity is the key for LLM as a Judge Anchor Selection
par: Don-Yehiya, Shachar, et autres
Publié: (2026)

When Can Transformers Count to n?
par: Yehudai, Gilad, et autres
Publié: (2024)

Preventing Rogue Agents Improves Multi-Agent Collaboration
par: Barbi, Ohav, et autres
Publié: (2025)

Routers Learn the Geometry of Their Experts: Geometric Coupling in Sparse Mixture-of-Experts
par: Ahrac, Sagi, et autres
Publié: (2026)

Mixing Mechanisms: How Language Models Retrieve Bound Entities In-Context
par: Gur-Arieh, Yoav, et autres
Publié: (2025)

Faithfulness Metrics Don't Measure Faithfulness: A Meta-Evaluation with Ground Truth
par: Gur-Arieh, Yoav, et autres
Publié: (2026)

Disentangling MLP Neuron Weights in Vocabulary Space
par: Avrahamy, Asaf, et autres
Publié: (2026)

Pretraining Language Models for Diachronic Linguistic Change Discovery
par: Fittschen, Elisabeth, et autres
Publié: (2025)

Holmes: A Benchmark to Assess the Linguistic Competence of Language Models
par: Waldis, Andreas, et autres
Publié: (2024)

ComPEFT: Compression for Communicating Parameter Efficient Updates via Sparsification and Quantization
par: Yadav, Prateek, et autres
Publié: (2023)

Deductive Closure Training of Language Models for Coherence, Accuracy, and Updatability
par: Akyürek, Afra Feyza, et autres
Publié: (2024)

Performance Gap in Entity Knowledge Extraction Across Modalities in Vision Language Models
par: Cohen, Ido, et autres
Publié: (2024)

Rethinking Selective Knowledge Distillation
par: Tavor, Almog, et autres
Publié: (2026)

Beneath the Surface of Consistency: Exploring Cross-lingual Knowledge Representation Sharing in LLMs
par: Ifergan, Maxim, et autres
Publié: (2024)

Will it Merge? On The Causes of Model Mergeability
par: Rahamim, Adir, et autres
Publié: (2026)

NumeroLogic: Number Encoding for Enhanced LLMs' Numerical Reasoning
par: Schwartz, Eli, et autres
Publié: (2024)

Indications of Belief-Guided Agency and Meta-Cognitive Monitoring in Large Language Models
par: Yalon, Noam Steinmetz, et autres
Publié: (2026)

Do LLMs Benefit From Their Own Words?
par: Huang, Jenny Y., et autres
Publié: (2026)

LLM Hypnosis: Exploiting User Feedback for Unauthorized Knowledge Injection to All Users
par: Hilel, Almog, et autres
Publié: (2025)

Friends and Grandmothers in Silico: Localizing Entity Cells in Language Models
par: Yona, Itay, et autres
Publié: (2026)