Stochastic Autoregressive Learning

• URL: http://arxiv.org/abs/2608.07224v1
• Date: Fri, 07 Aug 2026 13:40:17 GMT
• Title: Stochastic Autoregressive Learning
• Abstract summary: We introduce a PAC-learning model for binary autoregressive learning.<n>We show that autoregressive learning fundamentally differs from the deterministic theory.
• Score: 14.292982828097465
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Motivated by LLMs, which generate outputs by iteratively sampling from next-token distributions, we introduce a PAC-learning model for binary stochastic autoregressive learning. This generalizes the deterministic autoregressive learning framework of Joshi et al., COLT 2025. In our model, one fixed generator assigns a Bernoulli next-token distribution to every prompt string. Starting from an input prompt, a token is sampled and appended to the prompt; the same generator is then applied again to this expanded prompt; this procedure is repeated for $M$ steps. Three forms of supervision are considered: base one-step samples, chain-of-thought (CoT) samples that reveal full random trajectories of length $M$, and end-to-end (e2e) samples that reveal only the final token of length $M$ trajectories. For a generator class, we study the minimum number of samples $m_{base}(\varepsilon),m_{CoT}(\varepsilon), m_{e2e}(\varepsilon)$, resp., required to learn the one-step probabilities in the base model, and the final-token probability in the CoT and e2e models, under squared loss error~$\varepsilon$. We show that stochastic autoregressive learning fundamentally differs from the deterministic theory. At scale $\varepsilon$, there is no universal comparison between the three learning tasks: both $m_{CoT}/m_{base}$ and $m_{e2e}/m_{CoT}$ can be made simultaneously arbitrarily larger than $M/\varepsilon$, the natural analogue for the existing deterministic results. Nevertheless, after altering scales, for every class, CoT learning at scale $\varepsilon$ is upper-bounded by base learning at scale $\varepsilon/M^2$, whereas e2e learning at scale $\varepsilon$ is upper-bounded, up to logarithmic factors, by $(M/\varepsilon) m_{CoT}(Θ(\varepsilon))$. These dependencies and scales are essentially tight. We complement these bounds by studying dimension $d$ logistic functions in our model.

Related papers

• Efficient Sample-optimal Learning of Gaussian Tree Models via Sample-optimal Testing of Gaussian Mutual Information [1.7419682548187605]
  We develop a conditional mutual information tester for Gaussian random variables. We show that the chain rule of conditional mutual information continues to hold for the estimated (conditional) mutual information. We also show that when the underlying Gaussian model is not known to be tree-structured, $widetildeTheta(n2varepsilon-2)$ samples are necessary.
  arXiv  Detail & Related papers  (2024-11-18T12:25:34Z)
• Learning Orthogonal Multi-Index Models: A Fine-Grained Information Exponent Analysis [54.57279006229212]
  Information exponent has played an important role in predicting the sample complexity of online gradient descent.<n>In this work, we show that by considering both second- and higher-order terms, we can first learn the relevant space using the second-order terms.<n>The overall sample and complexity of online SGD is $tildeO( d PL-1 )$.
  arXiv  Detail & Related papers  (2024-10-13T00:14:08Z)
• Projection by Convolution: Optimal Sample Complexity for Reinforcement Learning in Continuous-Space MDPs [56.237917407785545]
  We consider the problem of learning an $varepsilon$-optimal policy in a general class of continuous-space Markov decision processes (MDPs) having smooth Bellman operators. Key to our solution is a novel projection technique based on ideas from harmonic analysis. Our result bridges the gap between two popular but conflicting perspectives on continuous-space MDPs.
  arXiv  Detail & Related papers  (2024-05-10T09:58:47Z)
• Approximation Rates and VC-Dimension Bounds for (P)ReLU MLP Mixture of Experts [17.022107735675046]
  Mixture-of-Experts (MoEs) can scale up beyond traditional deep learning models. We show that MoMLPs can generalize since the entire MoMLP model has a (finite) VC dimension of $tildeO(LmaxnL,JW)$.
  arXiv  Detail & Related papers  (2024-02-05T19:11:57Z)
• A Unified Framework for Uniform Signal Recovery in Nonlinear Generative Compressed Sensing [68.80803866919123]
  Under nonlinear measurements, most prior results are non-uniform, i.e., they hold with high probability for a fixed $mathbfx*$ rather than for all $mathbfx*$ simultaneously. Our framework accommodates GCS with 1-bit/uniformly quantized observations and single index models as canonical examples. We also develop a concentration inequality that produces tighter bounds for product processes whose index sets have low metric entropy.
  arXiv  Detail & Related papers  (2023-09-25T17:54:19Z)
• Replicability in Reinforcement Learning [46.89386344741442]
  We focus on the fundamental setting of discounted MDPs with access to a generative model. Inspired by Impagliazzo et al. [2022], we say that an RL algorithm is replicable if, with high probability, it outputs the exact same policy after two executions.
  arXiv  Detail & Related papers  (2023-05-31T05:16:23Z)
• Efficient Sampling of Stochastic Differential Equations with Positive Semi-Definite Models [91.22420505636006]
  This paper deals with the problem of efficient sampling from a differential equation, given the drift function and the diffusion matrix. It is possible to obtain independent and identically distributed (i.i.d.) samples at precision $varepsilon$ with a cost that is $m2 d log (1/varepsilon)$ Our results suggest that as the true solution gets smoother, we can circumvent the curse of dimensionality without requiring any sort of convexity.
  arXiv  Detail & Related papers  (2023-03-30T02:50:49Z)
• Tight Bounds on the Hardness of Learning Simple Nonparametric Mixtures [9.053430799456587]
  We study the problem of learning nonparametric distributions in a finite mixture. We establish tight bounds on the sample complexity for learning the component distributions in such models.
  arXiv  Detail & Related papers  (2022-03-28T23:53:48Z)
• Reward-Free Model-Based Reinforcement Learning with Linear Function Approximation [92.99933928528797]
  We study the model-based reward-free reinforcement learning with linear function approximation for episodic Markov decision processes (MDPs) In the planning phase, the agent is given a specific reward function and uses samples collected from the exploration phase to learn a good policy. We show that to obtain an $epsilon$-optimal policy for arbitrary reward function, UCRL-RFE needs to sample at most $tilde O(H4d(H + d)epsilon-2)$ episodes.
  arXiv  Detail & Related papers  (2021-10-12T23:03:58Z)
• Sample-Efficient Reinforcement Learning for Linearly-Parameterized MDPs with a Generative Model [3.749193647980305]
  This paper considers a Markov decision process (MDP) that admits a set of state-action features. We show that a model-based approach (resp.$$Q-learning) provably learns an $varepsilon$-optimal policy with high probability.
  arXiv  Detail & Related papers  (2021-05-28T17:49:39Z)
• Model-Free Reinforcement Learning: from Clipped Pseudo-Regret to Sample Complexity [59.34067736545355]
  Given an MDP with $S$ states, $A$ actions, the discount factor $gamma in (0,1)$, and an approximation threshold $epsilon > 0$, we provide a model-free algorithm to learn an $epsilon$-optimal policy. For small enough $epsilon$, we show an improved algorithm with sample complexity.
  arXiv  Detail & Related papers  (2020-06-06T13:34:41Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.