Cover Image

Detection vs. Execution: Single-Bucket Probes Miss Half the Mamba-2 State Sink

Fuente: arXiv

Saved in:

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

Similar Items

On the Existence and Behavior of Secondary Attention Sinks
by: Wong, Jeffrey T. H., et al.
Published: (2025)

Sink-Aware Pruning for Diffusion Language Models
by: Myrzakhan, Aidar, et al.
Published: (2026)

Stuffed Mamba: Oversized States Lead to the Inability to Forget
by: Chen, Yingfa, et al.
Published: (2024)

MambaQuant: Quantizing the Mamba Family with Variance Aligned Rotation Methods
by: Xu, Zukang, et al.
Published: (2025)

MemMamba: Rethinking Memory Patterns in State Space Model
by: Wang, Youjin, et al.
Published: (2025)

When Attention Sink Emerges in Language Models: An Empirical View
by: Gu, Xiangming, et al.
Published: (2024)

Attention Sinks: A 'Catch, Tag, Release' Mechanism for Embeddings
by: Zhang, Stephen, et al.
Published: (2025)

Hidden State Poisoning Attacks against Mamba-based Language Models
by: Mercier, Alexandre Le, et al.
Published: (2026)

MoE-Mamba: Efficient Selective State Space Models with Mixture of Experts
by: Pióro, Maciej, et al.
Published: (2024)

Differential Mamba
by: Schneider, Nadav, et al.
Published: (2025)

OrthoRank: Token Selection via Sink Token Orthogonality for Efficient LLM inference
by: Shin, Seungjun, et al.
Published: (2025)

The Computational Limits of State-Space Models and Mamba via the Lens of Circuit Complexity
by: Chen, Yifang, et al.
Published: (2024)

Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs
by: Kossen, Jannik, et al.
Published: (2024)

Exploring the Limitations of Mamba in COPY and CoT Reasoning
by: Ren, Ruifeng, et al.
Published: (2024)

NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Model
by: NVIDIA, et al.
Published: (2025)

Synthetic vs. Gold: The Role of LLM Generated Labels and Data in Cyberbullying Detection
by: Kazemi, Arefeh, et al.
Published: (2025)

Rhetorical Questions in LLM Representations: A Linear Probing Study
by: Yao, Louie Hong, et al.
Published: (2026)

Max It or Miss It: Benchmarking LLM On Solving Extremal Problems
by: Gao, Binxin, et al.
Published: (2025)

Post-Trained MoE Can Skip Half Experts via Self-Distillation
by: Lv, Xingtai, et al.
Published: (2026)

When Numbers Tell Half the Story: Human-Metric Alignment in Topic Model Evaluation
by: Prouteau, Thibault, et al.
Published: (2026)

Towards Execution-Grounded Automated AI Research
by: Si, Chenglei, et al.
Published: (2026)

From Emergence to Control: Probing and Modulating Self-Reflection in Language Models
by: Zhu, Xudong, et al.
Published: (2025)

BlackMamba: Mixture of Experts for State-Space Models
by: Anthony, Quentin, et al.
Published: (2024)

Single-pass Detection of Jailbreaking Input in Large Language Models
by: Candogan, Leyla Naz, et al.
Published: (2025)

Accelerated AI Inference via Dynamic Execution Methods
by: Barad, Haim, et al.
Published: (2024)

Explore-Execute Chain: Towards an Efficient Structured Reasoning Paradigm
by: Yang, Kaisen, et al.
Published: (2025)

Nemotron-H: A Family of Accurate and Efficient Hybrid Mamba-Transformer Models
by: NVIDIA, et al.
Published: (2025)

Missing Premise exacerbates Overthinking: Are Reasoning Models losing Critical Thinking Skill?
by: Fan, Chenrui, et al.
Published: (2025)

R2Gen-Mamba: A Selective State Space Model for Radiology Report Generation
by: Sun, Yongheng, et al.
Published: (2024)

Probe Pruning: Accelerating LLMs through Dynamic Pruning via Model-Probing
by: Le, Qi, et al.
Published: (2025)

Lost in Execution: On the Multilingual Robustness of Tool Calling in Large Language Models
by: Luo, Zheng, et al.
Published: (2026)

TIER: Trajectory-Invariant Execution Rewards for Multi-Step Tool Composition
by: Kulkarni, Anay, et al.
Published: (2026)

Flaming-hot Initiation with Regular Execution Sampling for Large Language Models
by: Chen, Weizhe, et al.
Published: (2024)

ASPERA: A Simulated Environment to Evaluate Planning for Complex Action Execution
by: Coca, Alexandru, et al.
Published: (2025)

Executable Functional Abstractions: Inferring Generative Programs for Advanced Math Problems
by: Khan, Zaid, et al.
Published: (2025)

In Search of Needles in a 11M Haystack: Recurrent Memory Finds What LLMs Miss
by: Kuratov, Yuri, et al.
Published: (2024)

Nemotron 3 Super: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning
by: NVIDIA, et al.
Published: (2026)

Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning
by: NVIDIA, et al.
Published: (2025)

The Information Geometry of Softmax: Probing and Steering
by: Park, Kiho, et al.
Published: (2026)

Building Production-Ready Probes For Gemini
by: Kramár, János, et al.
Published: (2026)