arXiv subjects

Andrew Chi-Chih Yao

Publications and source records attributed to Andrew Chi-Chih Yao.

At least 19 recordsLinked to original sources

arXiv2026

Fast Weight Attention for Continual Learning

Recurrent fast-weight memories and selective state-space models compress an expanding context into a fixed-size recurrent state, making the state transition an online learning rule. We study this rule under read-after-write autoregressive semantics. For the prefix-prediction objective considered here, the local fast-memory example revealed at step $t$ is the prefix-aligned pair $(\mathbf{x}_t,\mathbf{y}_t)=(ϕ(\mathbf{k}_{t-1}),\mathbf{v}_t)$. The common same-step association $(ϕ(\mathbf{k}_t),\mathbf{v}_t)$ remains causal, but optimizes a different internal objective. We derive normalized first-order updates for squared-error regression and negative inner-product objectives. The regression family comprises Falcon-1 (a scalar NLMS update), Falcon-2 (its per-column extension), and Falcon-3 (a sliding-window mini-batch update); Falcon-1A/Falcon-2A/Falcon-3A are the corresponding inner-product variants. We provide recurrent, masked-parallel, and chunk-parallel forms, together with numerically stable positive-decay renormalization. Representative variants remain competitive in language modeling and improve length extrapolation on variable-digit addition. This framework separates temporal alignment, plasticity, forgetting, and bounded rehearsal in recurrent sequence models.

cs.LG↗

arXiv2026

Meta Prompting for AI Systems

We introduce Meta Prompting (MP), a framework that emphasizes the formal structure of a task rather than content-specific worked examples. We give a categorical formalization in which a functor maps typed task transformations to typed prompt transformations. Functoriality encodes preservation of identities and composition, but it does not by itself guarantee semantic correctness. We extend MP to Recursive Meta Prompting (RMP), in which a proposer model generates candidate prompt edits, a validator enforces the edit schema, and an executor model uses the resulting prompt. We model the accumulation of valid edit scripts with the Writer monad and their application by a monoid action on prompts. This construction makes accumulated edits independent of parenthesization; it does not guarantee convergence or improved task accuracy. Empirically, a Qwen-72B base model guided by example-free meta-prompts attains 46.3% pass@1 on MATH and 83.5% accuracy on GSM8K. Separately, an MP-CR agent synthesizes a tool-executed Game of 24 solver; the original experiment reports 100% success on 1,362 instances, subject to exact re-verification from the corresponding input and per-instance outputs. The batch amortizes one LLM response across all puzzles, so its token accounting is not protocol-matched to per-instance few-shot or search-based methods.

cs.AI↗

arXiv2026

Monadic Context Engineering

The proliferation of Large Language Models (LLMs) has catalyzed a shift towards autonomous agents capable of complex reasoning and tool use. However, current agent architectures are frequently constructed using imperative, ad hoc patterns. This results in brittle systems plagued by difficulties in state management, error handling, and concurrency. This paper introduces Monadic Context Engineering (MCE), a novel architectural paradigm leveraging the algebraic structures of Functors, Applicative Functors, and Monads to provide a formal foundation for agent design. MCE treats agent workflows as computational contexts where cross-cutting concerns, such as state propagation, short-circuiting error handling, and asynchronous execution, are managed intrinsically by the algebraic properties of the abstraction. We demonstrate how Monads enable robust sequential composition, how Applicatives provide a principled structure for parallel execution, and crucially, how Monad Transformers allow for the systematic composition of these capabilities. This layered approach enables developers to construct complex, resilient, and efficient AI agents from simple, independently verifiable components. We further extend this framework to describe Meta-Agents, which leverage MCE for generative orchestration, dynamically creating and managing sub-agent workflows through metaprogramming.

cs.AI↗

arXiv2026

Cumulative Reasoning with Large Language Models

Recent advancements in large language models (LLMs) have shown remarkable progress, yet their ability to solve complex problems remains limited. In this work, we introduce Cumulative Reasoning (CR), a structured framework that enhances LLM problem-solving by emulating human-like iterative and cumulative thought processes. CR orchestrates LLMs in three distinct roles: Proposer, Verifier(s), and Reporter, to systematically decompose tasks, generate and validate intermediate reasoning steps, and compose them into a solution by building a dynamic Directed Acyclic Graph (DAG) of verified propositions. This approach substantially enhances problem-solving capabilities. We demonstrate CR's advantage through several complex reasoning tasks: it outperforms existing methods in logical inference tasks with up to a 9.3% improvement, achieving 98.04% accuracy on the curated FOLIO wiki dataset. In the Game of 24, it achieves 98% accuracy, marking a 24% improvement over previous methods. In solving MATH problems, CR achieves a 4.2% increase from previous methods and a 43% relative improvement in the most challenging level 5 problems. When incorporating a code environment with CR, we further harness LLMs' reasoning capabilities and outperform the Program of Thought (PoT) method by 38.8%.

cs.AI↗

arXiv2026

On the Diagram of Thought

Large Language Models (LLMs) excel at many tasks but often falter on complex problems that require structured, multi-step reasoning. We introduce the Diagram of Thought (DoT), a framework that enables a single LLM to build and navigate a mental map of its reasoning. Instead of thinking in a straight line, the model constructs a dynamic diagram of ideas, where it can propose different lines of thought, critique its own steps, and synthesize validated insights into a final conclusion. This process is controller-light: it does not require an external search algorithm or planner, but it does use a deterministic online validator for grammar-constrained typed traces, register constraints, and optional solver checks. To clarify the reliability target of this process, we ground DoT in a mathematical framework from category theory. We interpret accepted typed reasoning records as diagrams in a slice topos and model synthesis of the selected proposer subdiagram as a finite limit. In the predicate fragment, this same object is equivalently a variance-reversed colimit in the opposite information order. The resulting formalism gives an auditable, step-by-step trace of the LLM's typed reasoning and separates semantic guarantees for the typed subtrace from unconstrained natural-language text and uncertified operational edges.

cs.CL↗

arXiv2026

Group Representational Position Encoding

We present GRAPE (Group Representational Position Encoding), a unified framework for positional encoding based on group actions. GRAPE unifies two families of mechanisms: (i) multiplicative rotations (Multiplicative GRAPE) in $\operatorname{SO}(d)$ and (ii) additive logit biases (Additive GRAPE) arising from unipotent actions in the general linear group $\mathrm{GL}$. In Multiplicative GRAPE, a position $n \in \mathbb{Z}$ (or $t \in \mathbb{R}$) acts as $\mathbf{G}(n) = \exp(n \, ω\, \mathbf{L})$ with a rank-2 skew-symmetric generator $\mathbf{L} \in \mathbb{R}^{d \times d}$, yielding a relative, compositional, norm-preserving map with a closed-form matrix exponential. RoPE is recovered exactly when the $d/2$ planes correspond to canonical coordinate pairs with a log-uniform spectrum. Learned commuting subspaces and compact non-commuting mixtures strictly extend this geometry to capture cross-subspace feature coupling at $O(d)$ and $O(r d)$ cost per head, respectively. In Additive GRAPE, additive logits arise from rank-1 (or low-rank) unipotent actions, recovering ALiBi and the Forgetting Transformer (FoX) as exact special cases while preserving an exact relative law and streaming cacheability. Overall, GRAPE provides a principled design space for positional geometry in long-context models, subsuming RoPE and ALiBi as special cases. Project page: https://github.com/model-architectures/GRAPE.

cs.LG↗

arXiv2026

Probing the Lack of Stable Internal Beliefs in LLMs

Persona-driven large language models (LLMs) require consistent behavioral tendencies across interactions to simulate human-like personality traits, such as persistence or reliability. However, current LLMs often lack stable internal representations that anchor their responses over extended dialogues. This work explores whether LLMs can maintain "implicit consistency", defined as persistent adherence to an unstated goal in multi-turn interactions. We designed a 20-question-style riddle game paradigm where an LLM is tasked with secretly selecting a target and responding to users' guesses with "yes/no" answers. Through evaluations, we find that LLMs struggle to preserve latent consistency: their implicit "goals" shift across turns unless explicitly provided their selected target in context. These findings highlight critical limitations in the building of persona-driven LLMs and underscore the need for mechanisms that anchor implicit goals over time, which is a key to realistic personality modeling in interactive applications such as dialogue systems.

cs.CL↗

arXiv2026

On the Design of KL-Regularized Policy Gradient Algorithms for LLM Reasoning

Policy gradient algorithms have been successfully applied to enhance the reasoning capabilities of large language models (LLMs). KL regularization is ubiquitous, yet the design surface, choice of KL direction (forward vs. reverse), normalization (normalized vs. unnormalized), and estimator ($k_1/k_2/k_3$), is scattered across the literature and often intertwined with off-policy estimation. We ask a focused question: under the off-policy setting, what weighting is required for each KL variant so that the surrogate we optimize yields the exact gradient of the intended KL-regularized objective? We answer this with a compact, unified derivation we call the Regularized Policy Gradient (RPG) view. RPG (i) unifies normalized and unnormalized KL variants and shows that the widely-used $k_3$ penalty is exactly the unnormalized KL; (ii) specifies conditions under which REINFORCE-style losses with stop-gradient are gradient-equivalent to fully differentiable surrogates; (iii) identifies and corrects an off-policy importance-weighting mismatch in GRPO's KL term; and (iv) introduces RPG-Style Clip, a clipped-importance-sampling step within RPG-REINFORCE that enables stable, off-policy policy-gradient training at scale. On mathematical reasoning benchmarks (AIME24, AIME25), RPG-REINFORCE with RPG-Style Clip improves accuracy by up to $+6$ absolute percentage points over DAPO. We extend our experiments to 8K context length, and RPG-REINFORCE with RPG-Style Clip achieves 52% accuracy on AIME25, surpassing the official Qwen3-4B-Instruct model (47%). Notably, RPG is a stable and scalable RL algorithm for LLM reasoning, realized via (a) a KL-correct objective, (b) clipped importance sampling, and (c) an iterative reference-policy update scheme. Project Page: https://github.com/complex-reasoning/RPG.

cs.LG↗

arXiv2026

Tensor Product Attention Is All You Need

Scaling language models to handle longer input sequences typically necessitates large key-value (KV) caches, resulting in substantial memory overhead during inference. In this paper, we propose Tensor Product Attention (TPA), a novel attention mechanism that uses tensor decompositions to represent queries, keys, and values compactly, substantially shrinking the KV cache size at inference time. By factorizing these representations into contextual low-rank components and seamlessly integrating with Rotary Position Embedding (RoPE), TPA achieves improved model quality alongside memory efficiency. Based on TPA, we introduce the Tensor ProducT ATTenTion Transformer (T6), a new model architecture for sequence modeling. Through extensive empirical evaluation on language modeling tasks, we demonstrate that T6 surpasses or matches the performance of standard Transformer baselines including Multi-Head Attention (MHA), Multi-Query Attention (MQA), Grouped-Query Attention (GQA), and Multi-Head Latent Attention (MLA) across various metrics, including perplexity and a range of established evaluation benchmarks. Notably, TPA's memory efficiency and computational efficiency at decoding stage enables processing longer sequences under fixed resource constraints, addressing a critical scalability challenge in modern language models. Project Page: https://github.com/tensorgi/TPA.

cs.CL↗

arXiv2025

Clarifying Before Reasoning: A Coq Prover with Structural Context

In this work, we investigate whether improving task clarity can enhance reasoning ability of large language models, focusing on theorem proving in Coq. We introduce a concept-level metric to evaluate task clarity and show that adding structured semantic context to the standard input used by modern LLMs, leads to a 1.85$\times$ improvement in clarity score (44.5\%~$\rightarrow$~82.3\%). Using the general-purpose model \texttt{DeepSeek-V3}, our approach leads to a 2.1$\times$ improvement in proof success (21.8\%~$\rightarrow$~45.8\%) and outperforms the previous state-of-the-art \texttt{Graph2Tac} (33.2\%). We evaluate this on 1,386 theorems randomly sampled from 15 standard Coq packages, following the same evaluation protocol as \texttt{Graph2Tac}. Furthermore, fine-tuning smaller models on our structured data can achieve even higher performance (48.6\%). Our method uses selective concept unfolding to enrich task descriptions, and employs a Planner--Executor architecture. These findings highlight the value of structured task representations in bridging the gap between understanding and reasoning.

cs.AI↗

arXiv2025

Existing LLMs Are Not Self-Consistent For Simple Tasks

Large Language Models (LLMs) have grown increasingly powerful, yet ensuring their decisions remain transparent and trustworthy requires self-consistency -- no contradictions in their internal reasoning. Our study reveals that even on simple tasks, such as comparing points on a line or a plane, or reasoning in a family tree, all smaller models are highly inconsistent, and even state-of-the-art models like DeepSeek-R1 and GPT-o4-mini are not fully self-consistent. To quantify and mitigate these inconsistencies, we introduce inconsistency metrics and propose two automated methods -- a graph-based and an energy-based approach. While these fixes provide partial improvements, they also highlight the complexity and importance of self-consistency in building more reliable and interpretable AI. The code and data are available at https://github.com/scorpio-nova/llm-self-consistency.

cs.CL↗

arXiv2025

Towards Automated Formal Verification of Backend Systems with LLMs

Software testing plays a critical role in ensuring that systems behave as intended. However, existing automated testing approaches struggle to match the capabilities of human engineers due to key limitations such as test locality, lack of general reliability, and business logic blindness. In this work, we propose a novel framework that leverages functional programming and type systems to translate Scala backend code into formal Lean representations. Our pipeline automatically generates theorems that specify the intended behavior of APIs and database operations, and uses LLM-based provers to verify them. When a theorem is proved, the corresponding logic is guaranteed to be correct and no further testing is needed. If the negation of a theorem is proved instead, it confirms a bug. In cases where neither can be proved, human intervention is required. We evaluate our method on realistic backend systems and find that it can formally verify over 50% of the test requirements, which suggests that half of a testing engineer's workload can be automated. Additionally, with an average cost of only $2.19 per API, LLM-based verification is significantly more cost-effective than manual testing and can be scaled easily through parallel execution. Our results indicate a promising direction for scalable, AI-powered software testing, with the potential to greatly improve engineering productivity as models continue to advance.

cs.SE↗

arXiv2024

Augmenting Math Word Problems via Iterative Question Composing

Despite the advancements in large language models (LLMs) for mathematical reasoning, solving competition-level math problems remains a significant challenge, especially for open-source LLMs without external tools. We introduce the MMIQC dataset, comprising a mixture of processed web data and synthetic question-response pairs, aimed at enhancing the mathematical reasoning capabilities of base language models. Models fine-tuned on MMIQC consistently surpass their counterparts in performance on the MATH benchmark across various model sizes. Notably, Qwen-72B-MMIQC achieves a 45.0% accuracy, exceeding the previous open-source state-of-the-art by 8.2% and outperforming the initial version GPT-4 released in 2023. Extensive evaluation results on Hungarian high school finals suggest that such improvement can generalize to unseen data. Our ablation study on MMIQC reveals that a large part of the improvement can be attributed to our novel augmentation method, Iterative Question Composing (IQC), which involves iteratively composing new questions from seed problems using an LLM and applying rejection sampling through another LLM.

cs.CL↗

arXiv2023

Towards Data-Algorithm Dependent Generalization: a Case Study on Overparameterized Linear Regression

One of the major open problems in machine learning is to characterize generalization in the overparameterized regime, where most traditional generalization bounds become inconsistent even for overparameterized linear regression. In many scenarios, this failure can be attributed to obscuring the crucial interplay between the training algorithm and the underlying data distribution. This paper demonstrate that the generalization behavior of overparameterized model should be analyzed in a both data-relevant and algorithm-relevant manner. To make a formal characterization, We introduce a notion called data-algorithm compatibility, which considers the generalization behavior of the entire data-dependent training trajectory, instead of traditional last-iterate analysis. We validate our claim by studying the setting of solving overparameterized linear regression with gradient descent. Specifically, we perform a data-dependent trajectory analysis and derive a sufficient condition for compatibility in such a setting. Our theoretical results demonstrate that if we take early stopping iterates into consideration, generalization can hold with significantly weaker restrictions on the problem instance than the previous last-iterate analysis.

cs.LG↗

arXiv2021

FedCM: Federated Learning with Client-level Momentum

Federated Learning is a distributed machine learning approach which enables model training without data sharing. In this paper, we propose a new federated learning algorithm, Federated Averaging with Client-level Momentum (FedCM), to tackle problems of partial participation and client heterogeneity in real-world federated learning applications. FedCM aggregates global gradient information in previous communication rounds and modifies client gradient descent with a momentum-like term, which can effectively correct the bias and improve the stability of local SGD. We provide theoretical analysis to highlight the benefits of FedCM. We also perform extensive empirical studies and demonstrate that FedCM achieves superior performance in various tasks and is robust to different levels of client numbers, participation rate and client heterogeneity.

cs.LG↗

arXiv2020

An authenticated and secure accounting system for international emissions trading

Expanding multi-country emissions trading system is considered as crucial to fill the existing mitigation gap for the 2\degree C climate target. Trustworthy emissions accounting is the cornerstone of such a system encompassing different jurisdictions. However, traditional emissions measuring, reporting, and verification practices that support data authenticity might not be applicable as detailed data from large utilities and production facilities to be covered in the multi-country emissions trading system are usually highly sensitive and of severe national security concern. In this study, we propose a cryptographic framework for an authenticated and secure emissions accounting system that can resolve this data dilemma. We demonstrate that integrating a sequence of cryptographic protocols can preserve data authenticity and security for a stylized multi-country emissions trading system. We call for more research to promote applications of modern cryptography in future international climate governance to build trust and strengthen collaboration.

econ.GN↗

arXiv2018

An Incentive Analysis of some Bitcoin Fee Designs

In the Bitcoin system, miners are incentivized to join the system and validate transactions through fees paid by the users. A simple "pay your bid" auction has been employed to determine the transaction fees. Recently, Lavi, Sattath and Zohar [LSZ17] proposed an alternative fee design, called the monopolistic price (MP) mechanism, aimed at improving the revenue for the miners. Although MP is not strictly incentive compatible (IC), they studied how close to IC the mechanism is for iid distributions, and conjectured that it is nearly IC asymptotically based on extensive simulations and some analysis. In this paper, we prove that the MP mechanism is nearly incentive compatible for any iid distribution as the number of users grows large. This holds true with respect to other attacks such as splitting bids. We also prove a conjecture in [LSZ17] that MP dominates the RSOP auction in revenue (originally defined in Goldberg et al. [GHKSW06] for digital goods). These results lend support to MP as a Bitcoin fee design candidate. Additionally, we explore some possible intrinsic correlations between incentive compatibility and revenue in general.

cs.GT↗

arXiv2017

On Revenue Monotonicity in Combinatorial Auctions

Along with substantial progress made recently in designing near-optimal mechanisms for multi-item auctions, interesting structural questions have also been raised and studied. In particular, is it true that the seller can always extract more revenue from a market where the buyers value the items higher than another market? In this paper we obtain such a revenue monotonicity result in a general setting. Precisely, consider the revenue-maximizing combinatorial auction for $m$ items and $n$ buyers in the Bayesian setting, specified by a valuation function $v$ and a set $F$ of $nm$ independent item-type distributions. Let $REV(v, F)$ denote the maximum revenue achievable under $F$ by any incentive compatible mechanism. Intuitively, one would expect that $REV(v, G)\geq REV(v, F)$ if distribution $G$ stochastically dominates $F$. Surprisingly, Hart and Reny (2012) showed that this is not always true even for the simple case when $v$ is additive. A natural question arises: Are these deviations contained within bounds? To what extent may the monotonicity intuition still be valid? We present an {approximate monotonicity} theorem for the class of fractionally subadditive (XOS) valuation functions $v$, showing that $REV(v, G)\geq c\,REV(v, F)$ if $G$ stochastically dominates $F$ under $v$ where $c>0$ is a universal constant. Previously, approximate monotonicity was known only for the case $n=1$: Babaioff et al. (2014) for the class of additive valuations, and Rubinstein and Weinberg (2015) for all subaddtive valuation functions.

cs.GT↗