Copertina

Do LLMs "know" internally when they follow instructions?

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Heo, Juyeon, Heinze-Deml, Christina, Elachqar, Oussama, Chan, Kwan Ho Ryan, Ren, Shirley, Nallasamy, Udhay, Miller, Andy, Narain, Jaya
Natura:	Preprint
Pubblicazione:	2024
Soggetti:	Artificial Intelligence Computation and Language
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Do LLMs estimate uncertainty well in instruction-following?
di: Heo, Juyeon, et al.
Pubblicazione: (2024)

Large-scale Training of Foundation Models for Wearable Biosignals
di: Abbaspourazad, Salar, et al.
Pubblicazione: (2023)

Speech Foundation Models Generalize to Time Series Tasks from Wearable Sensor Data
di: Narain, Jaya, et al.
Pubblicazione: (2025)

Towards Provably Unbiased LLM Judges via Bias-Bounded Evaluation
di: Feuer, Benjamin, et al.
Pubblicazione: (2026)

Characterization and Greedy Learning of Gaussian Structural Causal Models under Unknown Interventions
di: Gamella, Juan L., et al.
Pubblicazione: (2022)

MARVIS: Modality Adaptive Reasoning over VISualizations
di: Feuer, Benjamin, et al.
Pubblicazione: (2025)

ButterflyQuant: Ultra-low-bit LLM Quantization through Learnable Orthogonal Butterfly Transforms
di: Xu, Bingxin, et al.
Pubblicazione: (2025)

Considerations for Distribution Shift Robustness of Diagnostic Models in Healthcare
di: Blaas, Arno, et al.
Pubblicazione: (2024)

Affect Models Have Weak Generalizability to Atypical Speech
di: Narain, Jaya, et al.
Pubblicazione: (2025)

Anti-causal domain generalization: Leveraging unlabeled data
di: Saengkyongam, Sorawit, et al.
Pubblicazione: (2026)

Do Concept Bottleneck Models Respect Localities?
di: Raman, Naveen, et al.
Pubblicazione: (2024)

Hybrid Modeling of Photoplethysmography for Non-invasive Monitoring of Cardiovascular Parameters
di: Palumbo, Emanuele, et al.
Pubblicazione: (2025)

On Evaluating LLMs' Capabilities as Functional Approximators: A Bayesian Perspective
di: Siddiqui, Shoaib Ahmed, et al.
Pubblicazione: (2024)

Using LLMs for Late Multimodal Sensor Fusion for Activity Recognition
di: Demirel, Ilker, et al.
Pubblicazione: (2025)

When Judgment Becomes Noise: How Design Failures in LLM Judge Benchmarks Silently Undermine Validity
di: Feuer, Benjamin, et al.
Pubblicazione: (2025)

KDA: A Knowledge-Distilled Attacker for Generating Diverse Prompts to Jailbreak LLMs
di: Liang, Buyun, et al.
Pubblicazione: (2025)

Artificial or Just Artful? Do LLMs Bend the Rules in Programming?
di: Sghaier, Oussama Ben, et al.
Pubblicazione: (2025)

Ferrocene Appended Linear Chromophores for Aggregation‐Induced Emission (AIE) and Nonlinear Optics (NLO): Combined Experimental and Theoretical Studies
di: Vadakkalur Sampath Chithra, et al.
Pubblicazione: (2024)

AIE‐Active Cyano Substituted Dimethoxy–Phenyl Derivatives for Nonlinear Optics: Spectral, Structural, and DFT Studies
di: Vadakkalur Sampath Chithra, et al.
Pubblicazione: (2025)

Unveiling Legitimacy in the unexpected events context : An Inquiry into Information System Consultancy companies and international organizations through Topic Modeling Analysis
di: Abidi, Oussama
Pubblicazione: (2024)

Contextual Knowledge Pursuit for Faithful Visual Synthesis
di: Luo, Jinqi, et al.
Pubblicazione: (2023)

IP-CRR: Information Pursuit for Interpretable Classification of Chest Radiology Reports
di: Ge, Yuyan, et al.
Pubblicazione: (2025)

Leveraging Task Structures for Improved Identifiability in Neural Network Representations
di: Chen, Wenlin, et al.
Pubblicazione: (2023)

Advancing biomolecular understanding and design following human instructions
di: Zhuang, Xiang, et al.
Pubblicazione: (2024)

IWISDM: Assessing instruction following in multimodal models at scale
di: Lei, Xiaoxuan, et al.
Pubblicazione: (2024)

El federalismo de la India está plagado de altercados por las vías fluviales
di: Narain, A
Pubblicazione: (2009)

Water Security, Conflict and Cooperation in Peri-Urban South Asia Flows across Boundaries
di: Vishal Narain

Statistical genomics and bioinformatics
di: Prem Narain
Pubblicazione: (2010)

Allium jammuense (Amaryllidaceae: Allieae) a New Species From Subgen. Cepa (Mill.) Prokh. From Trikuta Hills of Jammu and Kashmir, India
di: Sunit Singh, et al.
Pubblicazione: (2026)

MRI for axial SpA: Diagnosis, disease activity assessment, and recent advances
di: Shirley Chiu Wai Chan, et al.
Pubblicazione: (2024)

Effects of pH on the Flavor Detection Threshold and DoT of Nine Kokumi Peptides in Soybean Fermented Foods
di: Juyeon Lee, et al.
Pubblicazione: (2025)

Estimation of Concept Explanations Should be Uncertainty Aware
di: Piratla, Vihari, et al.
Pubblicazione: (2023)

BLEUBERI: BLEU is a surprisingly effective reward for instruction following
di: Chang, Yapei, et al.
Pubblicazione: (2025)

Oral submucosal fibrosis: An updated molecular mechanism on pathogenesis and treatment modalities
di: Vadivel Jayanth Kumar, et al.
Pubblicazione: (2024)

Probing Embodied LLMs: When Higher Observation Fidelity Hurts Problem Solving
di: Zenkri, Oussama, et al.
Pubblicazione: (2026)

Do you know what q-means?
di: Cornelissen, Arjan, et al.
Pubblicazione: (2023)

‘I don't know my way about’: Mirror reversal as a curiously instructive analogue of philosophical perplexity
di: Andrew English
Pubblicazione: (2024)

Conformal Information Pursuit for Interactively Guiding Large Language Models
di: Chan, Kwan Ho Ryan, et al.
Pubblicazione: (2025)

Can a Single Model Master Both Multi-turn Conversations and Tool Use? CoALM: A Unified Conversational Agentic Language Model
di: Acikgoz, Emre Can, et al.
Pubblicazione: (2025)

Enabling robots to follow abstract instructions and complete complex dynamic tasks
di: Mon-Williams, Ruaridh, et al.
Pubblicazione: (2024)