Cover Image

When Slower Isn't Truer: Inverse Scaling Law of Truthfulness in Multimodal Reasoning

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Fang, Sitong, Cao, Wenjing, Li, Jiahao, Wang, Xuyao, Dai, Juntao, Chan, Chi-Min, Han, Sirui, Guo, Yike, Yang, Yaodong, Ji, Jiaming
Format:	Preprint
Published:	2025
Subjects:	Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Inverse Scaling: When Bigger Isn't Better
by: McKenzie, Ian R., et al.
Published: (2023)

Reasoning Isn't Enough: Examining Truth-Bias and Sycophancy in LLMs
by: Barkett, Emilio, et al.
Published: (2025)

ThinkPatterns-21k: A Systematic Study on the Impact of Thinking Patterns in LLMs
by: Wen, Pengcheng, et al.
Published: (2025)

When Trust Isn't Enough.
by: Behrman, Sara
Published: (1998)

SafeMT: Multi-turn Safety for Multimodal Language Models
by: Zhu, Han, et al.
Published: (2025)

SafeLawBench: Towards Safe Alignment of Large Language Models
by: Cao, Chuxue, et al.
Published: (2025)

Mitigating Deceptive Alignment via Self-Monitoring
by: Ji, Jiaming, et al.
Published: (2025)

Thinking Isn't an Illusion: Overcoming the Limitations of Reasoning Models via Tool Augmentations
by: Song, Zhao, et al.
Published: (2025)

When Fairness Isn't Statistical: The Limits of Machine Learning in Evaluating Legal Reasoning
by: Barale, Claire, et al.
Published: (2025)

If It Isn't Broken...Break It!
by: Voges, Mickie A.
Published: (2000)

The Universe Isn't Expanding, It's Relaxing
by: Blouin, Sam
Published: (2025)

The Universe Isn't Expanding — It's Relaxing
by: Blouin, Sam
Published: (2025)

AI Isn't Creating Anything
by: Lee Skallerup Bessette
Published: (2025)

Videodiscs: A Revolution That Isn't.
Published: (1982)

AI Isn't Creating Anything
by: Lee Skallerup Bessette
Published: (2025)

Climate Adaptation Money Isn't Reaching the Most Vulnerable— And Why It Matters
by: Đại Bàng
Published: (2025)

When Alignment Isn't Enough: Response-Path Attacks on LLM Agents
by: Luo, Mingyu, et al.
Published: (2026)

When More Isn't Better: The Curvilinear Effects of ESG on Firm Performance
by: Joel Victor Dossa, et al.
Published: (2026)

J1: Exploring Simple Test-Time Scaling for LLM-as-a-Judge
by: Chan, Chi-Min, et al.
Published: (2025)

ML Interpretability: Simple Isn't Easy
by: Räz, Tim
Published: (2022)

"Just Say No" Isn't Sex Education.
by: Osborn, Anne
Published: (1991)

When Meaning Isn't Literal: Exploring Idiomatic Meaning Across Languages and Modalities
by: Das, Sarmistha, et al.
Published: (2026)

When Privacy Isn't Synthetic: Hidden Data Leakage in Generative AI Models
by: Mustaqim, S. M., et al.
Published: (2025)

When Standard Newborn Screening Isn't Enough: Diagnostic Challenges in the Age of Globalization
by: Marina Ortúzar Menéndez, et al.
Published: (2026)

Awake ECMO for Mid‐Tracheal Obstruction: When a Tracheostomy Isn't Enough
by: Jacob Beiriger, et al.
Published: (2026)

SafeMCP: Proactive Power Regulation for LLM Agent Defense via Environment-Grounded Look-Ahead Reasoning
by: Wang, Lichao, et al.
Published: (2026)

Knowing the Answer Isn't Enough: Fixing Reasoning Path Failures in LVLMs
by: Wang, Chaoyang, et al.
Published: (2025)

Strong Reasoning Isn't Enough: Evaluating Evidence Elicitation in Interactive Diagnosis
by: Long, Zhuohan, et al.
Published: (2026)

What, Whether and How? Unveiling Process Reward Models for Thinking with Images Reasoning
by: Zhou, Yujin, et al.
Published: (2026)

The Newberys: Getting Them Read (It Isn't Easy)
by: Aborne, Carlene
Published: (1974)

The Future Isn't What It Used to Be: Videotex Is on the Way.
by: McKenzie, Jamieson A.
Published: (1984)

Excuse Me, Isn't That Your Library on Fire?
by: Grayson, Randall
Published: (1998)

Take the Step—Waiting Isn’t a Strategy
by: David B. LaFrance
Published: (2026)

When One Moment Isn't Enough: Multi-Moment Retrieval with Cross-Moment Interactions
by: Cao, Zhuo, et al.
Published: (2025)

When Correct Isn't Usable: Improving Structured Output Reliability in Small Language Models
by: Galeone, Cosimo, et al.
Published: (2026)

When 'For You' Isn't For You: Measuring User Agency in TikTok's Algorithmic Feed
by: Kaplan, Levi, et al.
Published: (2026)

Seeing Isn't Knowing: Do VLMs Know When Not to Answer Spatial Questions (and Why)?
by: Zhang, Yue, et al.
Published: (2026)

When Kids Mode Isn't For Kids: Investigating TikTok's "Under 13 Experience"
by: Figueira, Olivia, et al.
Published: (2025)

Comment on “Awake ECMO for Mid‐Tracheal Obstruction: When a Tracheostomy Isn't Enough”
by: Ilaria Onorati, et al.
Published: (2026)

InterMT: Multi-Turn Interleaved Preference Alignment with Human Feedback
by: Chen, Boyuan, et al.
Published: (2025)