Copertina

The Illusion of Diminishing Returns: Measuring Long Horizon Execution in LLMs

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Sinha, Akshit, Arun, Arvindh, Goel, Shashwat, Staab, Steffen, Geiping, Jonas
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Artificial Intelligence
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

FutureSim: Replaying World Events to Evaluate Adaptive Agents
di: Goel, Shashwat, et al.
Pubblicazione: (2026)

A Cognac Shot To Forget Bad Memories: Corrective Unlearning for Graph Neural Networks
di: Kolipaka, Varshita, et al.
Pubblicazione: (2024)

The Illusion of Procedural Reasoning: Measuring Long-Horizon FSM Execution in LLMs
di: Samiei, Mahdi, et al.
Pubblicazione: (2025)

Pitfalls in Evaluating Language Model Forecasters
di: Paleka, Daniel, et al.
Pubblicazione: (2025)

Answer Matching Outperforms Multiple Choice for Language Model Evaluation
di: Chandak, Nikhil, et al.
Pubblicazione: (2025)

SEMMA: A Semantic Aware Knowledge Graph Foundation Model
di: Arun, Arvindh, et al.
Pubblicazione: (2025)

The Diminishing Returns of Early-Exit Decoding in Modern LLMs
di: Wei, Rui, et al.
Pubblicazione: (2026)

Intrinsic Credit Assignment for Long Horizon Interaction
di: Auzina, Ilze Amanda, et al.
Pubblicazione: (2026)

On the Diminishing Returns of Width for Continual Learning
di: Guha, Etash, et al.
Pubblicazione: (2024)

LLMs4Life: Large Language Models for Ontology Learning in Life Sciences
di: Fathallah, Nadeen, et al.
Pubblicazione: (2024)

Self-Consistency Is Losing Its Edge: Diminishing Returns and Rising Costs in Modern LLMs
di: Loo, Chiyan
Pubblicazione: (2025)

AccessGuru: Leveraging LLMs to Detect and Correct Web Accessibility Violations in HTML Code
di: Fathallah, Nadeen, et al.
Pubblicazione: (2025)

Fine, I'll Merge It Myself: A Multi-Fidelity Framework for Automated Model Merging
di: Su, Guinan, et al.
Pubblicazione: (2025)

EcoGym: Evaluating LLMs for Long-Horizon Plan-and-Execute in Interactive Economies
di: Hu, Xavier, et al.
Pubblicazione: (2026)

From Tokens to Lattices: Emergent Lattice Structures in Language Models
di: Xiong, Bo, et al.
Pubblicazione: (2025)

Efficiently Dispatching Flash Attention For Partially Filled Attention Masks
di: Sharma, Agniv, et al.
Pubblicazione: (2024)

Great Models Think Alike and this Undermines AI Oversight
di: Goel, Shashwat, et al.
Pubblicazione: (2025)

Empowering the Deaf and Hard of Hearing Community: Enhancing Video Captions Using Large Language Models
di: Fathallah, Nadeen, et al.
Pubblicazione: (2024)

CAFIN: Centrality Aware Fairness inducing IN-processing for Unsupervised Representation Learning on Graphs
di: Arun, Arvindh, et al.
Pubblicazione: (2023)

Intrinsic Stability Limits of Autoregressive Reasoning: Structural Consequences for Long-Horizon Execution
di: Liao, Hsien-Jyh
Pubblicazione: (2026)

The Semantic Illusion: Certified Limits of Embedding-Based Hallucination Detection in RAG Systems
di: Sinha, Debu
Pubblicazione: (2025)

PARC: An Autonomous Self-Reflective Coding Agent for Robust Execution of Long-Horizon Tasks
di: Orimo, Yuki, et al.
Pubblicazione: (2025)

SAGE: Scene Graph-Aware Guidance and Execution for Long-Horizon Manipulation Tasks
di: Li, Jialiang, et al.
Pubblicazione: (2025)

F -- A Model of Events based on the Foundational Ontology DOLCE+DnS Ultralite
di: Scherp, Ansgar, et al.
Pubblicazione: (2024)

Training High-Level Schedulers with Execution-Feedback Reinforcement Learning for Long-Horizon GUI Automation
di: Deng, Zehao, et al.
Pubblicazione: (2025)

When Robots Do the Chores: A Benchmark and Agent for Long-Horizon Household Task Execution
di: Zhu, Zilin, et al.
Pubblicazione: (2026)

Cooperative Evolutionary Pressure and Diminishing Returns Might Explain the Fermi Paradox: On What Super-AIs Are Like
di: Vallstrom, Daniel
Pubblicazione: (2024)

Strict Subgoal Execution: Reliable Long-Horizon Planning in Hierarchical Reinforcement Learning
di: Hwang, Jaebak, et al.
Pubblicazione: (2025)

Can LLMs Introspect? A Reality Check
di: Singh, Shashwat, et al.
Pubblicazione: (2026)

Expanding Expressivity in Transformer Models with MöbiusAttention
di: Halacheva, Anna-Maria, et al.
Pubblicazione: (2024)

Can you Finetune your Binoculars? Embedding Text Watermarks into the Weights of Large Language Models
di: Elhassan, Fay, et al.
Pubblicazione: (2025)

Models That Know How Evaluations Are Designed Score Safer
di: Deckenbach, Katharina, et al.
Pubblicazione: (2026)

Security Risks in Tool-Enabled AI Agents: A Systematic Analysis of Privileged Execution Environments
di: Goel, Hardik
Pubblicazione: (2026)

Performance of LLMs on Stochastic Modeling Operations Research Problems: From Theory to Practice
di: Kumar, Akshit, et al.
Pubblicazione: (2025)

The Self-Execution Benchmark: Measuring LLMs' Attempts to Overcome Their Lack of Self-Execution
di: Ezra, Elon, et al.
Pubblicazione: (2025)

Technical Debt in In-Context Learning: Diminishing Efficiency in Long Context
di: Joo, Taejong, et al.
Pubblicazione: (2025)

Can Language Models Falsify? Evaluating Algorithmic Reasoning with Counterexample Creation
di: Sinha, Shiven, et al.
Pubblicazione: (2025)

When Fewer Layers Break More Chains: Layer Pruning Harms Test-Time Scaling in LLMs
di: Wang, Keyu, et al.
Pubblicazione: (2025)

SELFDOUBT: Uncertainty Quantification for Reasoning LLMs via the Hedge-to-Verify Ratio
di: Pandey, Satwik, et al.
Pubblicazione: (2026)

The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution
di: Li, Junlong, et al.
Pubblicazione: (2025)