Image de couverture de livre

Deep sequence models tend to memorize geometrically; it is unclear why

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Noroozizadeh, Shahriar, Nagarajan, Vaishnavh, Rosenfeld, Elan, Kumar, Sanjiv
Format:	Preprint
Publié:	2025
Sujets:	Machine Learning Artificial Intelligence 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

The pitfalls of next-token prediction
par: Bachmann, Gregor, et autres
Publié: (2024)

Think before you speak: Training Language Models With Pause Tokens
par: Goyal, Sachin, et autres
Publié: (2023)

Reconstructing Sepsis Trajectories from Clinical Case Reports using LLMs: the Textual Time Series Corpus for Sepsis
par: Noroozizadeh, Shahriar, et autres
Publié: (2025)

TLDR at SemEval-2024 Task 2: T5-generated clinical-Language summaries for DeBERTa Report Analysis
par: Das, Spandan, et autres
Publié: (2024)

Text Knows What, Tables Know When: Clinical Timeline Reconstruction via Retrieval-Augmented Multimodal Alignment
par: Kumar, Sayantan, et autres
Publié: (2026)

PMOA-TTS: Introducing the PubMed Open Access Textual Times Series Corpus
par: Noroozizadeh, Shahriar, et autres
Publié: (2025)

Roll the dice & look before you leap: Going beyond the creative limits of next-token prediction
par: Nagarajan, Vaishnavh, et autres
Publié: (2025)

On student-teacher deviations in distillation: does it pay to disobey?
par: Nagarajan, Vaishnavh, et autres
Publié: (2023)

World Properties without World Models: Recovering Spatial and Temporal Structure from Co-occurrence Statistics in Static Word Embeddings
par: Barenholtz, Elan
Publié: (2026)

Forecasting Clinical Risk from Textual Time Series: Structuring Narratives for Temporal AI in Healthcare
par: Noroozizadeh, Shahriar, et autres
Publié: (2025)

ET tu, CLIP? Addressing Common Object Errors for Unseen Environments
par: Byun, Ye Won, et autres
Publié: (2024)

Reasoning with Latent Thoughts: On the Power of Looped Transformers
par: Saunshi, Nikunj, et autres
Publié: (2025)

Temporal Supervised Contrastive Learning for Modeling Patient Risk Progression
par: Noroozizadeh, Shahriar, et autres
Publié: (2023)

Data Contamination Quiz: A Tool to Detect and Estimate Contamination in Large Language Models
par: Golchin, Shahriar, et autres
Publié: (2023)

On the Inductive Bias of Stacking Towards Improving Reasoning
par: Saunshi, Nikunj, et autres
Publié: (2024)

Exploring and Improving Drafts in Blockwise Parallel Decoding
par: Kim, Taehyeon, et autres
Publié: (2024)

HU at SemEval-2024 Task 8A: Can Contrastive Learning Learn Embeddings to Detect Machine-Generated Text?
par: Dipta, Shubhashis Roy, et autres
Publié: (2024)

Language Model Cascades: Token-level uncertainty and beyond
par: Gupta, Neha, et autres
Publié: (2024)

Faster Cascades via Speculative Decoding
par: Narasimhan, Harikrishna, et autres
Publié: (2024)

Mechanistic origins of catastrophic forgetting: why RL preserves circuits better than SFT?
par: Nunez, Jeanmely Rojas, et autres
Publié: (2026)

Revisiting the Effectiveness of LLM Pruning for Test-Time Scaling
par: Monjur, Ocean, et autres
Publié: (2026)

The Impact of Medication Non-adherence on Adverse Outcomes: Evidence from Schizophrenia Patients via Survival Analysis
par: Noroozizadeh, Shahriar, et autres
Publié: (2025)

SurvHTE-Bench: A Benchmark for Heterogeneous Treatment Effect Estimation in Survival Analysis
par: Noroozizadeh, Shahriar, et autres
Publié: (2026)

LoRA Done RITE: Robust Invariant Transformation Equilibration for LoRA Optimization
par: Yen, Jui-Nan, et autres
Publié: (2024)

DistillSpec: Improving Speculative Decoding via Knowledge Distillation
par: Zhou, Yongchao, et autres
Publié: (2023)

Less Diverse, Less Safe: The Indirect But Pervasive Risk of Test-Time Scaling in Large Language Models
par: Nahin, Shahriar Kabir, et autres
Publié: (2025)

Time Travel in LLMs: Tracing Data Contamination in Large Language Models
par: Golchin, Shahriar, et autres
Publié: (2023)

Intent Laundering: AI Safety Datasets Are Not What They Seem
par: Golchin, Shahriar, et autres
Publié: (2026)

Memorization in In-Context Learning
par: Golchin, Shahriar, et autres
Publié: (2024)

Catch Your Breath: Adaptive Computation for Self-Paced Sequence Production
par: Galashov, Alexandre, et autres
Publié: (2025)

What are you sinking? A geometric approach on attention sink
par: Ruscio, Valeria, et autres
Publié: (2025)

Task Facet Learning: A Structured Approach to Prompt Optimization
par: Juneja, Gurusha, et autres
Publié: (2024)

Multi-Preference Optimization: Generalizing DPO via Set-Level Contrasts
par: Gupta, Taneesh, et autres
Publié: (2024)

DLBacktrace: A Model Agnostic Explainability for any Deep Learning Models
par: Sankarapu, Vinay Kumar, et autres
Publié: (2024)

Identifying Representations for Intervention Extrapolation
par: Saengkyongam, Sorawit, et autres
Publié: (2023)

MultiSoc-4D: A Benchmark for Diagnosing Instruction-Induced Label Collapse in Closed-Set LLM Annotation of Bengali Social Media
par: Pramanik, Souvik, et autres
Publié: (2026)

One-Shot Strategic Classification Under Unknown Costs
par: Rosenfeld, Elan, et autres
Publié: (2023)

PortLLM: Personalizing Evolving Large Language Models with Training-Free and Portable Model Patches
par: Khan, Rana Muhammad Shahroz, et autres
Publié: (2024)

Deep learning and abstractive summarisation for radiological reports: an empirical study for adapting the PEGASUS models' family with scarce data
par: Benzoni, Claudio, et autres
Publié: (2025)

Learning under noisy supervision is governed by a feedback-truth gap
par: Schonfeld, Elan, et autres
Publié: (2026)