paper-with-me

홈 › Papers

Rep2Text: Decoding Full Text from a Single LLM Token Representation

2025-11-09 · Haiyan Zhao, Zirui He, Yiming Tang, Fan Yang, Ali Payani, Dianbo Liu, Mengnan Du arxiv

Large language models (LLMs) have achieved remarkable progress across diverse tasks, yet their internal mechanisms remain largely opaque. In this work, we investigate a fundamental question: to what extent can the original input text be recovered from a single last-token representation in an LLM? To this end, we propose Rep2Text, a novel framework for decoding text from last-token representations. Rep2Text employs a trainable adapter that maps a target model's last-token representation into the token embedding space of a decoding language model, which then autoregressively reconstructs the input text. Experiments across various model combinations (Llama-3.1-8B, Gemma-7B, Mistral-7B-v0.1, Llama-3.2-3B, etc.) show that, on average, roughly half of the tokens in 16-token sequences can be recovered from this compressed representation while preserving strong semantic coherence. Further analysis reveals a clear information bottleneck effect: as sequence length increases, token-level recovery declines, while semantic information remains relatively well preserved. We also find that scaling effects are less pronounced in inversion tasks. Finally, our framework demonstrates robust generalization to out-of-distribution clinical data.

📄 PDF Abstract BibTeX arXiv:2511.06571

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Parallel Decoding via Hidden Transfer for Lossless Large Language Model Acceleration

2024-04-18 · Pengfei Wu, Jiahao Liu, Zhuocheng Gong, Qifan Wang 외

Large language models (LLMs) have recently shown remarkable performance across a wide range of tasks. However, the substantial number of parameters in LLMs contributes to significant latency during model inference. This …

Language ModelingLanguage ModellingLarge Language Model

$x$-Prediction Flow: Efficient Continuous Decoding for Masked Diffusion Language Models

2026-06-27 · Weitian Wang, Lianlei Shan, Shubham Rai, Cecilia De La Parra 외 arxiv

Masked diffusion language models (MDLMs) generate text by iteratively unmasking tokens, but their standard decoder reduces each step to a binary action: a position is either committed to a single token or left fully mask…

Reinforcement Learning

Stability-Weighted Decoding for Diffusion Language Models

2026-04-18 · Yue Wu, Jian Huang arxiv

Diffusion large language models (dLLMs) enable parallel text generation by iteratively denoising a fully masked sequence, unmasking a subset of masked tokens at each step. Existing decoding strategies rely on static conf…

Mathematical ReasoningCode GenerationText Generation

Finish First, Perfect Later: Test-Time Token-Level Cross-Validation for Diffusion Large Language Models

2025-10-06 · Runchu Tian, Junxia Cui, Xueqiang Xu, Feng Yao 외 arxiv

Diffusion large language models (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) models, offering advantages such as accelerated parallel decoding and bidirectional context modeling. Howeve…

Code Generation

SimSD: Simple Speculative Decoding in Diffusion Language Models

2026-06-01 · Junxia Cui, Haotian Ye, Runchu Tian, Hongcan Guo 외 arxiv

Diffusion large language models (dLLMs) have recently emerged as a promising alternative to autoregressive (AR) LLMs, offering faster inference through parallel or blockwise decoding. However, their masked language model…