Limitations on Autoregressive Token Generation via Sequential Sampling
Gal Arnon, Andrej Bogdanov, Paul Lou, Giulio Malavolta, Alon Rosen, Lorenzo Taschin
Abstract
We study autoregressive token generation through the notion of sequential samplers: A circuit that samples the next symbol only given the previous ones and possibly a small amount of state. We show that (1) a sequential sampler with k bits of state can be simulated by a stateless one at a roughly O(2k) size overhead; (2) efficient insertional sampling (Stern, Chan, Kiros, Uszkoreit, ICML 2019) is more powerful than efficient sequential sampling in any fixed order; (3) sequential samplers with small state cannot establish certain forms of covert cryptographic channels. Result (1), together with the pseudoentropy vs. sequential sampleability duality of Vadhan and Zheng (STOC 2012), characterizes efficient sequential sampleability with small state. Result (3) provides a potential safety mechanism against subliminal coordination among malicious LLM agents.
Create a lesson
Related papers
MSE Minimization for Movable Antenna Enhanced Wireless Sensing via Array Geometry Optimization
Wenrong Chen, Lipeng Zhu, Wenyan Ma et al.
Sum Rate Maximization for Practical RIS-Enhanced Rate-Splitting Multiple Access Systems
Qian Zhang, Xuejun Cheng, Jiaming Shi et al.
Optimal Receive Design on the Observation Axis for RSMA Satellite-UAV Downlinks
Xianwei Meng
DeepTelecom: A Differentiable Multimodal Framework for Channel Twinning in Dynamic Heterogeneous Wireless Environments
Bohao Wang, Tianen Yuan, William Yen et al.
DS and AMDS Symbol-Pair Codes from Three Families of Simple Root Cyclic Codes
Jiawei He, Zihan Zhang, Xiaoqiang Wang et al.
Low-Latency LDPC Decoding Based on Consensus and Return Residual Belief Propagation
Allan E. Feitosa, Rodrigo C. de Lamare