Copertina

Decomposing Representation Space into Interpretable Subspaces with Unsupervised Learning

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Huang, Xinting, Hahn, Michael
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Machine Learning Artificial Intelligence Computation and Language
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Discovering Interpretable Algorithms by Decompiling Transformers to RASP
di: Huang, Xinting, et al.
Pubblicazione: (2026)

InversionView: A General-Purpose Method for Reading Information from Neural Activations
di: Huang, Xinting, et al.
Pubblicazione: (2024)

Navigating Ideation Space: Decomposed Conceptual Representations for Positioning Scientific Ideas
di: Shen, Yuexi, et al.
Pubblicazione: (2026)

Understanding In-context Learning of Addition via Activation Subspaces
di: Hu, Xinyan, et al.
Pubblicazione: (2025)

Interpretable Safety Alignment via SAE-Constructed Low-Rank Subspace Adaptation
di: Wang, Dianyun, et al.
Pubblicazione: (2025)

Decomposing and Measuring Evaluation Awareness
di: Li, Changling, et al.
Pubblicazione: (2026)

DATA: Decomposed Attention-based Task Adaptation for Rehearsal-Free Continual Learning
di: Liao, Huanxuan, et al.
Pubblicazione: (2025)

SpreadsheetArena: Decomposing Preference in LLM Generation of Spreadsheet Workbooks
di: Kundurthy, Srivatsa, et al.
Pubblicazione: (2026)

Resolving Conflicts in Lifelong Learning via Aligning Updates in Subspaces
di: Zhou, Yueer, et al.
Pubblicazione: (2025)

Decomposing Attention To Find Context-Sensitive Neurons
di: Gibson, Alex
Pubblicazione: (2025)

SPARC: Subspace-Aware Prompt Adaptation for Robust Continual Learning in LLMs
di: Jayasuriya, Dinithi, et al.
Pubblicazione: (2025)

Contextualize-then-Aggregate: Circuits for In-Context Learning in Gemma-2 2B
di: Bakalova, Aleksandra, et al.
Pubblicazione: (2025)

Superscopes: Amplifying Internal Feature Representations for Language Model Interpretation
di: Jacobi, Jonathan, et al.
Pubblicazione: (2025)

Interpretable Discriminative Text Representations via Agreement and Label Disentanglement
di: Wang, Tong, et al.
Pubblicazione: (2026)

DecomposeRL: Learning to Ask Useful, Informative, and Diverse Questions for Semi-Supervised, Traceable Claim Verification
di: Dipta, Shubhashis Roy, et al.
Pubblicazione: (2026)

Selection-p: Self-Supervised Task-Agnostic Prompt Compression for Faithfulness and Transferability
di: Chung, Tsz Ting, et al.
Pubblicazione: (2024)

Mnemosyne: An Unsupervised, Human-Inspired Long-Term Memory Architecture for Edge-Based LLMs
di: Jonelagadda, Aneesh, et al.
Pubblicazione: (2025)

LLM Assertiveness can be Mechanistically Decomposed into Emotional and Logical Components
di: Tsujimura, Hikaru, et al.
Pubblicazione: (2025)

Learning Self-Interpretation from Interpretability Artifacts: Training Lightweight Adapters on Vector-Label Pairs
di: Pepper, Keenan, et al.
Pubblicazione: (2026)

Decomposing Elements of Problem Solving: What "Math" Does RL Teach?
di: Qin, Tian, et al.
Pubblicazione: (2025)

Efficient and Scalable Estimation of Tool Representations in Vector Space
di: Moon, Suhong, et al.
Pubblicazione: (2024)

Learning Unsupervised Semantic Document Representation for Fine-grained Aspect-based Sentiment Analysis
di: Fu, Hao-Ming, et al.
Pubblicazione: (2024)

Analytic Subspace Routing: How Recursive Least Squares Works in Continual Learning of Large Language Model
di: Tong, Kai, et al.
Pubblicazione: (2025)

Memory-Efficient LLM Training with Online Subspace Descent
di: Liang, Kaizhao, et al.
Pubblicazione: (2024)

ROSA: Random Subspace Adaptation for Efficient Fine-Tuning
di: Hameed, Marawan Gamal Abdel, et al.
Pubblicazione: (2024)

Too Long, Didn't Model: Decomposing LLM Long-Context Understanding With Novels
di: Hamilton, Sil, et al.
Pubblicazione: (2025)

PRISM: A Geometric Risk Bound that Decomposes Drift into Scale, Shape, and Head
di: Lin, Chieh-Yen, et al.
Pubblicazione: (2026)

Thoughtbubbles: an Unsupervised Method for Parallel Thinking in Latent Space
di: Liu, Houjun, et al.
Pubblicazione: (2025)

HyperDAS: Towards Automating Mechanistic Interpretability with Hypernetworks
di: Sun, Jiuding, et al.
Pubblicazione: (2025)

Training Acceleration of Low-Rank Decomposed Networks using Sequential Freezing and Rank Quantization
di: Hajimolahoseini, Habib, et al.
Pubblicazione: (2023)

Learning to Interpret Weight Differences in Language Models
di: Goel, Avichal, et al.
Pubblicazione: (2025)

EDoRA: Efficient Weight-Decomposed Low-Rank Adaptation via Singular Value Decomposition
di: Nasiri, Hamid, et al.
Pubblicazione: (2025)

Safety Subspaces are Not Linearly Distinct: A Fine-Tuning Case Study
di: Ponkshe, Kaustubh, et al.
Pubblicazione: (2025)

Unraveling LoRA Interference: Orthogonal Subspaces for Robust Model Merging
di: Zhang, Haobo, et al.
Pubblicazione: (2025)

LaMDA: Large Model Fine-Tuning via Spectrally Decomposed Low-Dimensional Adaptation
di: Azizi, Seyedarmin, et al.
Pubblicazione: (2024)

3DS: Medical Domain Adaptation of LLMs via Decomposed Difficulty-based Data Selection
di: Ding, Hongxin, et al.
Pubblicazione: (2024)

From Explainable to Interpretable Deep Learning for Natural Language Processing in Healthcare: How Far from Reality?
di: Huang, Guangming, et al.
Pubblicazione: (2024)

ICLR: In-Context Learning of Representations
di: Park, Core Francisco, et al.
Pubblicazione: (2024)

Decoding Diffusion: A Scalable Framework for Unsupervised Analysis of Latent Space Biases and Representations Using Natural Language Prompts
di: Zeng, E. Zhixuan, et al.
Pubblicazione: (2024)

LLM Self-Correction with DeCRIM: Decompose, Critique, and Refine for Enhanced Following of Instructions with Multiple Constraints
di: Ferraz, Thomas Palmeira, et al.
Pubblicazione: (2024)