Image de couverture de livre

Measuring and Controlling Instruction (In)Stability in Language Model Dialogs

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Li, Kenneth, Liu, Tianle, Bashkansky, Naomi, Bau, David, Viégas, Fernanda, Pfister, Hanspeter, Wattenberg, Martin
Format:	Preprint
Publié:	2024
Sujets:	Computation and Language Artificial Intelligence Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Task
par: Li, Kenneth, et autres
Publié: (2022)

Inference-Time Intervention: Eliciting Truthful Answers from a Language Model
par: Li, Kenneth, et autres
Publié: (2023)

Dialogue Action Tokens: Steering Language Models in Goal-Directed Dialogue with a Multi-Turn Planner
par: Li, Kenneth, et autres
Publié: (2024)

When Bad Data Leads to Good Models
par: Li, Kenneth, et autres
Publié: (2025)

Relational Composition in Neural Networks: A Survey and Call to Action
par: Wattenberg, Martin, et autres
Publié: (2024)

What Does it Mean for a Neural Network to Learn a "World Model"?
par: Li, Kenneth, et autres
Publié: (2025)

The Geometry of Self-Verification in a Task-Specific Reasoning Model
par: Lee, Andrew, et autres
Publié: (2025)

Can Interpretation Predict Behavior on Unseen Data?
par: Li, Victoria R., et autres
Publié: (2025)

Shared Global and Local Geometry of Language Model Embeddings
par: Lee, Andrew, et autres
Publié: (2025)

Discovering Forbidden Topics in Language Models
par: Rager, Can, et autres
Publié: (2025)

Measuring Progress in Dictionary Learning for Language Model Interpretability with Board Game Models
par: Karvonen, Adam, et autres
Publié: (2024)

LLäMmlein: Transparent, Compact and Competitive German-Only Language Models from Scratch
par: Pfister, Jan, et autres
Publié: (2024)

Towards a Zero-Data, Controllable, Adaptive Dialog System
par: Väth, Dirk, et autres
Publié: (2024)

Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Models
par: Marks, Samuel, et autres
Publié: (2024)

Large Language Models as Tool Makers
par: Cai, Tianle, et autres
Publié: (2023)

Tree of Attributes Prompt Learning for Vision-Language Models
par: Ding, Tong, et autres
Publié: (2024)

Why Can't Transformers Learn Multiplication? Reverse-Engineering Reveals Long-Range Dependency Pitfalls
par: Bai, Xiaoyan, et autres
Publié: (2025)

Learning from Relevant Subgoals in Successful Dialogs using Iterative Training for Task-oriented Dialog Systems
par: Kaiser, Magdalena, et autres
Publié: (2024)

From Language Modeling to Instruction Following: Understanding the Behavior Shift in LLMs after Instruction Tuning
par: Wu, Xuansheng, et autres
Publié: (2023)

Instruction Mining: Instruction Data Selection for Tuning Large Language Models
par: Cao, Yihan, et autres
Publié: (2023)

GPT-2 Through the Lens of Vector Symbolic Architectures
par: Knittel, Johannes, et autres
Publié: (2024)

Dialog2Flow: Pre-training Soft-Contrastive Action-Driven Sentence Embeddings for Automatic Dialog Flow Extraction
par: Burdisso, Sergio, et autres
Publié: (2024)

Optimizing Instructions and Demonstrations for Multi-Stage Language Model Programs
par: Opsahl-Ong, Krista, et autres
Publié: (2024)

Large Language Models Generate Harmful Content Using a Distinct, Unified Mechanism
par: Orgad, Hadas, et autres
Publié: (2026)

Communicating Activations Between Language Model Agents
par: Ramesh, Vignav, et autres
Publié: (2025)

Instruction-tuned Language Models are Better Knowledge Learners
par: Jiang, Zhengbao, et autres
Publié: (2024)

Instruction Tuning for Large Language Models: A Survey
par: Zhang, Shengyu, et autres
Publié: (2023)

Nevermind: Instruction Override and Moderation in Large Language Models
par: Kim, Edward
Publié: (2024)

What Causes Polysemanticity? An Alternative Origin Story of Mixed Selectivity from Incidental Causes
par: Lecomte, Victor, et autres
Publié: (2023)

Conifer: Improving Complex Constrained Instruction-Following Ability of Large Language Models
par: Sun, Haoran, et autres
Publié: (2024)

Small Language Models: Survey, Measurements, and Insights
par: Lu, Zhenyan, et autres
Publié: (2024)

An Empirical Study on Context Length for Open-Domain Dialog Generation
par: Shen, Xinyi, et autres
Publié: (2024)

Conversational Tree Search: A New Hybrid Dialog Task
par: Väth, Dirk, et autres
Publié: (2023)

CaLM: Contrasting Large and Small Language Models to Verify Grounded Generation
par: Hsu, I-Hung, et autres
Publié: (2024)

CodecLM: Aligning Language Models with Tailored Synthetic Data
par: Wang, Zifeng, et autres
Publié: (2024)

Improving Instruction-Following in Language Models through Activation Steering
par: Stolfo, Alessandro, et autres
Publié: (2024)

Federated Data-Efficient Instruction Tuning for Large Language Models
par: Qin, Zhen, et autres
Publié: (2024)

SFTMix: Elevating Language Model Instruction Tuning with Mixup Recipe
par: Xiao, Yuxin, et autres
Publié: (2024)

Instruction Following by Principled Boosting Attention of Large Language Models
par: Guardieiro, Vitoria, et autres
Publié: (2025)

Reward Collapse in Aligning Large Language Models
par: Song, Ziang, et autres
Publié: (2023)