Cover Image

Climbing the Ladder of Reasoning: What LLMs Can-and Still Can't-Solve after SFT?

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Sun, Yiyou, Zhou, Georgia, Bai, Haoyue, Wang, Hao, Li, Dacheng, Dziri, Nouha, Song, Dawn
Format:	Preprint
Published:	2025
Subjects:	Artificial Intelligence Computation and Language Machine Learning
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

OMEGA: Can LLMs Reason Outside the Box in Math? Evaluating Exploratory, Compositional, and Transformative Generalization
by: Sun, Yiyou, et al.
Published: (2025)

RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?
by: Sun, Yiyou, et al.
Published: (2025)

Can LLMs Ask Good Questions?
by: Zhang, Yueheng, et al.
Published: (2025)

LLMs Still Can't Plan; Can LRMs? A Preliminary Evaluation of OpenAI's o1 on PlanBench
by: Valmeekam, Karthik, et al.
Published: (2024)

LLMs Can't Handle Peer Pressure: Crumbling under Multi-Agent Social Interactions
by: Song, Maojia, et al.
Published: (2025)

Strategy Executability in Mathematical Reasoning: Leveraging Human-Model Differences for Effective Guidance
by: Liang, Weida, et al.
Published: (2026)

Can't say cant? Measuring and Reasoning of Dark Jargons in Large Language Models
by: Ji, Xu, et al.
Published: (2024)

Surfacing Semantic Orthogonality Across Model Safety Benchmarks: A Multi-Dimensional Analysis
by: Bennion, Jonathan, et al.
Published: (2025)

TurnWise: The Gap between Single- and Multi-turn Language Model Capabilities
by: Graf, Victoria, et al.
Published: (2026)

You Can't Steal Nothing: Mitigating Prompt Leakages in LLMs via System Vectors
by: Cao, Bochuan, et al.
Published: (2025)

We Can't Understand AI Using our Existing Vocabulary
by: Hewitt, John, et al.
Published: (2025)

Can't See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs
by: Wang, Wenxuan, et al.
Published: (2025)

Catch Me If You Can? Not Yet: LLMs Still Struggle to Imitate the Implicit Writing Styles of Everyday Authors
by: Wang, Zhengxiang, et al.
Published: (2025)

Even GPT-5.2 Can't Count to Five: The Case for Zero-Error Horizons in Trustworthy LLMs
by: Sato, Ryoma
Published: (2026)

Your Teacher Can't Help You Here: Combating Supervision Fidelity Decay in On-Policy Distillation
by: Liu, Yanjiang, et al.
Published: (2026)

Rel-A.I.: An Interaction-Centered Approach To Measuring Human-LM Reliance
by: Zhou, Kaitlyn, et al.
Published: (2024)

Can Reasoning LLMs Enhance Clinical Document Classification?
by: Mustafa, Akram, et al.
Published: (2025)

WildBench: Benchmarking LLMs with Challenging Tasks from Real Users in the Wild
by: Lin, Bill Yuchen, et al.
Published: (2024)

Beyond I'm Sorry, I Can't: Dissecting Large Language Model Refusal
by: Prakash, Nirmalendu, et al.
Published: (2025)

Can Language Models Solve Olympiad Programming?
by: Shi, Quan, et al.
Published: (2024)

AbsenceBench: Language Models Can't Tell What's Missing
by: Fu, Harvey Yiyun, et al.
Published: (2025)

The Generative AI Paradox on Evaluation: What It Can Solve, It May Not Evaluate
by: Oh, Juhyun, et al.
Published: (2024)

If You Can't Use Them, Recycle Them: Optimizing Merging at Scale Mitigates Performance Tradeoffs
by: Khalifa, Muhammad, et al.
Published: (2024)

Why and How LLMs Hallucinate: Connecting the Dots with Subsequence Associations
by: Sun, Yiyou, et al.
Published: (2025)

A Simple "Motivation" Can Enhance Reinforcement Finetuning of Large Reasoning Models
by: Zhang, Junjie, et al.
Published: (2025)

Can Post-Training Transform LLMs into Causal Reasoners?
by: Chen, Junqi, et al.
Published: (2026)

Empathy and the Right to Be an Exception: What LLMs Can and Cannot Do
by: Kidder, William, et al.
Published: (2024)

Phenomenal Yet Puzzling: Testing Inductive Reasoning Capabilities of Language Models with Hypothesis Refinement
by: Qiu, Linlu, et al.
Published: (2023)

Semantic Deception: When Reasoning Models Can't Compute an Addition
by: de Leeuw, Nathaniël, et al.
Published: (2025)

Can RL Teach Long-Horizon Reasoning to LLMs? Expressiveness Is Key
by: Wang, Tianle, et al.
Published: (2026)

Can LLMs Learn by Teaching for Better Reasoning? A Preliminary Study
by: Ning, Xuefei, et al.
Published: (2024)

Johnny Still Can't Read
by: Melcher, Daniel
Published: (1973)

Evaluate What You Can't Evaluate: Unassessable Quality for Generated Response
by: Liu, Yongkang, et al.
Published: (2023)

LLMs Can't Plan, But Can Help Planning in LLM-Modulo Frameworks
by: Kambhampati, Subbarao, et al.
Published: (2024)

Can LLMs Correct Themselves? A Benchmark of Self-Correction in LLMs
by: Tie, Guiyao, et al.
Published: (2025)

Can't Remember Details in Long Documents? You Need Some R&R
by: Agrawal, Devanshu, et al.
Published: (2024)

Frontier LLMs Still Struggle with Simple Reasoning Tasks
by: Malek, Alan, et al.
Published: (2025)

Quagmires in SFT-RL Post-Training: When High SFT Scores Mislead and What to Use Instead
by: Kang, Feiyang, et al.
Published: (2025)

What Can String Probability Tell Us About Grammaticality?
by: Hu, Jennifer, et al.
Published: (2025)

Random Initialization Can't Catch Up: The Advantage of Language Model Transfer for Time Series Forecasting
by: Riachi, Roland, et al.
Published: (2025)