Cover Image

Adapter Merging Reactivates Latent Reasoning Traces: A Mechanism Analysis

Fuente: arXiv

Saved in:

Bibliographic Details
Main Author:	Zou, Junyi
Format:	Preprint
Published:	2026
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Instruction-Tuned, but Not More Verifiable Instruction-Following: A Cross-Task Diagnosis for LoRA Adapters
by: Zou, Junyi
Published: (2026)

AdaMergeX: Cross-Lingual Transfer with Large Language Models via Adaptive Adapter Merging
by: Zhao, Yiran, et al.
Published: (2024)

ReasonOps: Operator Segmentation for LLM Reasoning Traces
by: Lee, Daniel, et al.
Published: (2026)

K-Merge: Online Continual Merging of Adapters for On-device Large Language Models
by: Shenaj, Donald, et al.
Published: (2025)

Reasoning Pattern Alignment Merging for Adaptive Reasoning
by: Zhong, Zhaofeng, et al.
Published: (2026)

Data-driven Clustering and Merging of Adapters for On-device Large Language Models
by: Bohdal, Ondrej, et al.
Published: (2026)

Reasoning Models Hallucinate More: Factuality-Aware Reinforcement Learning for Large Reasoning Models
by: Li, Junyi, et al.
Published: (2025)

DAM: Dynamic Adapter Merging for Continual Video QA Learning
by: Cheng, Feng, et al.
Published: (2024)

RCP-Merging: Merging Long Chain-of-Thought Models with Domain-Specific Models by Considering Reasoning Capability as Prior
by: Yang, Junyao, et al.
Published: (2025)

Compressing Sequences in the Latent Embedding Space: $K$-Token Merging for Large Language Models
by: Xu, Zihao, et al.
Published: (2026)

Multi-objective Evolutionary Merging Enables Efficient Reasoning Models
by: Iacobelli, Mario, et al.
Published: (2026)

ReasonAny: Incorporating Reasoning Capability to Any Model via Simple and Effective Model Merging
by: Yang, Junyao, et al.
Published: (2026)

ReasoningShield: Safety Detection over Reasoning Traces of Large Reasoning Models
by: Li, Changyi, et al.
Published: (2025)

Latent Reasoning with Supervised Thinking States
by: Amos, Ido, et al.
Published: (2026)

Latent Chain-of-Thought for Visual Reasoning
by: Sun, Guohao, et al.
Published: (2025)

MedAdapter: Efficient Test-Time Adaptation of Large Language Models towards Medical Reasoning
by: Shi, Wenqi, et al.
Published: (2024)

LaRS: Latent Reasoning Skills for Chain-of-Thought Reasoning
by: Xu, Zifan, et al.
Published: (2023)

Revisiting the UID Hypothesis in LLM Reasoning Traces
by: Gwak, Minju, et al.
Published: (2025)

TRAIL: Trace Reasoning and Agentic Issue Localization
by: Deshpande, Darshan, et al.
Published: (2025)

The Thinking Spectrum: An Empirical Study of Tunable Reasoning in LLMs through Model Merging
by: Lan, Xiaochong, et al.
Published: (2025)

Finding and Reactivating Post-Trained LLMs' Hidden Safety Mechanisms
by: Li, Mingjie, et al.
Published: (2026)

Extrapolation Merging: Keep Improving With Extrapolation and Merging
by: Lin, Yiguan, et al.
Published: (2025)

SE-Merging: A Self-Enhanced Approach for Dynamic Model Merging
by: Chen, Zijun, et al.
Published: (2025)

What Makes Good Multilingual Reasoning? Disentangling Reasoning Traces with Measurable Features
by: Ki, Dayeon, et al.
Published: (2026)

Emergent Search and Backtracking in Latent Reasoning Models
by: Cui, Jasmine, et al.
Published: (2026)

Latent Reasoning via Sentence Embedding Prediction
by: Hwang, Hyeonbin, et al.
Published: (2025)

Learning to Ponder: Adaptive Reasoning in Latent Space
by: He, Yixin, et al.
Published: (2025)

SuperMerge: An Approach For Gradient-Based Model Merging
by: Yang, Haoyu, et al.
Published: (2024)

SwiReasoning: Switch-Thinking in Latent and Explicit for Pareto-Superior Reasoning LLMs
by: Shi, Dachuan, et al.
Published: (2025)

Filtered Reasoning Score: Evaluating Reasoning Quality on a Model's Most-Confident Traces
by: Pathak, Manas, et al.
Published: (2026)

Transport and Merge: Cross-Architecture Merging for Large Language Models
by: Cui, Chenhang, et al.
Published: (2026)

Comparative Analysis of Efficient Adapter-Based Fine-Tuning of State-of-the-Art Transformer Models
by: Siddiqui, Saad Mashkoor, et al.
Published: (2025)

Lessons from Studying Two-Hop Latent Reasoning
by: Balesni, Mikita, et al.
Published: (2024)

Beyond Words: A Latent Memory Approach to Internal Reasoning in LLMs
by: Orlicki, José I.
Published: (2025)

TraceSIR: A Multi-Agent Framework for Structured Analysis and Reporting of Agentic Execution Traces
by: Yang, Shu-Xun, et al.
Published: (2026)

InfoDensity: Rewarding Information-Dense Traces for Efficient Reasoning
by: Wei, Chengwei, et al.
Published: (2026)

Evaluating Legal Reasoning Traces with Legal Issue Tree Rubrics
by: Lee, Jinu, et al.
Published: (2025)

Do Cognitively Interpretable Reasoning Traces Improve LLM Performance?
by: Bhambri, Siddhant, et al.
Published: (2025)

Can GNN be Good Adapter for LLMs?
by: Huang, Xuanwen, et al.
Published: (2024)

Parameter-Efficient Fine-Tuning With Adapters
by: Chen, Keyu, et al.
Published: (2024)