arXiv subjects

Mo Li

Publications and source records attributed to Mo Li.

At least 19 recordsLinked to original sources

arXiv2026

Estimating GHG Emissions from AI Use: Framework for Corporate-Level Measurement

Electricity demand from data centers is expected to grow from roughly 5% of U.S. consumption in 2025 to between 9% and 17% by 2030, and corporate artificial intelligence (AI) use is following a similar trajectory, spanning employee productivity assistants, direct access to large language models (LLMs), and AI features embedded in enterprise software. AI emissions today are a small share of footprints for many enterprises, but that share is unlikely to remain small for long. Without reasonable estimates, companies cannot set reduction targets or identify effective decarbonization levers as emissions grow. Companies, regulators, and auditors are asking for emissions estimates that withstand scrutiny, but no widely accepted methodology exists today. Published per-query estimates can differ by several orders of magnitude depending on what is counted, which provider is measured, and what assumptions are made about electricity use and the grid mix. This white paper proposes a standardized framework for corporate-level AI emissions accounting. The framework is designed to be defensible with current data constraints, tiered to meet companies where their data are, transparent about its assumptions, updatable as provider disclosure matures, and built for action rather than disclosure alone. Since AI emissions accounting is still nascent, it has the opportunity to design for actionability from the outset, so that measurement incentivizes responsible choices during AI's rapid buildout.

physics.soc-ph↗

arXiv2026

Training Skills Like Parameters via Self-Supervised Semantic Diffusion

While Large Language Models (LLMs) demonstrate remarkable general instruction-following capabilities, they often fall short of human experts in highly specialized, open-ended domains such as creative screenwriting. Prior approaches typically adopt post-training, yet both supervised fine-tuning and reinforcement learning require weight access that closed-source frontier models do not offer, and demand heavy compute. Moreover, what is learned is tied to a single checkpoint and cannot be inspected by humans. Recent advancements in agentic continual learning instead attempt to bridge this gap by accumulating external textual skills. However, these methods heavily rely on costly human expert annotations or unreliable LLM-as-a-judge feedback for reflection. To overcome this bottleneck, we propose a novel, unsupervised self-evolving agent framework inspired by the corruption-and-reconstruction paradigm of diffusion models. Instead of relying on explicit external scoring, we leverage existing high-quality human artifacts to construct self-supervised signals. Training then follows the familiar loop of neural network training, forward, loss, and backward, with the loss coming from contrasting the agent's reconstruction against the human original. What is updated is not model weights but an external library of textual skills. We evaluate our framework on the challenging task of short drama screenwriting. Experimental results demonstrate that our method enables the agent to autonomously extract and internalize highly generalizable skills, significantly enhancing its domain-specific generation capabilities. Furthermore, this self-contrastive reflection paradigm offers a scalable pathway for agents to teach themselves the production of complex, high-quality human artifacts, without requiring external supervision.

cs.CL↗

arXiv2026

LLM-Based Generative Retrieval for Snapchat Content Recommendation

Pretrained large language models (LLMs) are promising retrieval engines because they combine rich semantic priors, strong sequence modeling capabilities, and favorable scaling behavior. However, turning a pretrained LLM into a generative retriever in production deployment raises several challenges: the model must learn an internal item vocabulary that was absent from pretraining, and generate valid item identifiers under strict latency and cost constraints. We address these challenges through the design and launch of SnapLGR, an LLM-based generative retrieval system for short-video recommendation at Snapchat. The system is built around three main designs. First, we construct semantic identifiers (SIDs) from multimodal item embeddings and enhance them with Personalized PageRank (PPR)-based co-engagement contrastive learning, resulting in improved codebook utilization, reduced collisions, and infused collaborative signal. Second, we use continued pretraining (CPT) to ground the introduced SID tokens before supervised fine-tuning (SFT) on user interaction sequences. Third, we make SnapLGR serving practical through TensorRT-LLM CUDA-backed beam search and a decentralized worker-loop architecture. In a live A/B test, the launched system increased View Time by 0.37%, Time Spent by 0.09%, Deep Sessions by 0.18%, and Deep Sessions Unique User by 0.11% relative to the existing TIGER-style generative retrieval baseline. We then decompose this offline gap under a fixed tokenizer and quantify the gains due to model architecture, scaling, and pretraining. Overall, our deployment shows that successful production SnapLGR requires joint design across representation learning, vocabulary grounding, and efficient training and serving.

cs.IR↗

arXiv2026

MemTX: Transactional Belief Commit for Stateful Agent Memory

LLM agents increasingly coordinate through persistent shared memory: one agent's write becomes another agent's premise, and eventually a tool call with real side effects. Current agent memory systems treat every accepted write as immediately actionable truth, so a polluted tool result, a stale update, or a teammate's half-finished note can silently drive an irreversible action. We argue that a memory write is not a belief commit. We present MemTX, a transactional belief-commit protocol. Each record carries evidence, permissions, provenance, and validity. Writes are staged inside snapshot-isolated transactions and admitted by a validate-and-commit pipeline, irreversible tool calls are gated on in-flight belief state, and retracting a belief triggers typed cascading repair of its derived records and tool side effects. Two invariants, action-safety gating and cascade-repair completeness, are machine-checked by property-based testing and bounded exhaustive enumeration of 5.5 million protocol states, with zero violations. Across five backbones from three model families, MemTX leads all eight baselines with paired-McNemar significance on four backbones and statistically ties the best baseline on the fifth and strongest, while remaining the only method with zero downstream harm on every backbone. Backbone capability does not substitute for commit discipline.

cs.AI↗

arXiv2026

Strategic Plan for Neutral Atom Quantum Computation

We present a strategic plan for neutral atom quantum computation, bringing together hardware development and theory advancements to achieve the goal of practical quantum advantage. The concept of practical quantum advantage is defined, along with how to verify claims of advantage, and approaches to designing quantum algorithms that deliver practical advantage. Future directions for neutral atom quantum processor hardware are described: scaling-up system size, Qubit encodings and atomic platforms, going further below threshold with neutral-atom logical-qubit performance, continuous reloading of qubits, and fast readout. We also explore opportunities for scalable integrated photonic control technologies. Alongside hardware advancements, new developments in quantum error correction and compilation of quantum circuits are proposed. Finally, we examine the opportunity of networking multiple neutral atom quantum processors together to perform distributed quantum computing and overcome possible limitations of a single system.

quant-ph↗

arXiv2026

AgentCompass: A Unified Evaluation Infrastructure for Agent Capabilities

As Large Language Models (LLMs) evolve into autonomous agents, the need for unified evaluation infrastructure becomes critical. However, current evaluation pipelines remain highly fragmented and tightly coupled, hindering reproducibility and causing redundant engineering. To address this, we introduce AgentCompass, an open-source, lightweight, and extensible infrastructure for evaluating LLM-based agents. AgentCompass organizes the evaluation process around three independent components, namely Benchmark, Harness, and Environment, thereby enabling flexible configurations without requiring the reimplementation of complex execution logic. Furthermore, it features a fault-tolerant asynchronous runtime and comprehensive trajectory analysis tools to transparently diagnose nuanced failure modes like reward-hacking. Natively supporting over 20 benchmarks across five capability dimensions, AgentCompass provides the community with a scalable and reproducible infrastructure for advancing agent research.

cs.AI↗

arXiv2026

GatedLinear: Adaptive Routing of Complementary Linear Bases for Time Series Forecasting

Time series forecasting requires models to capture diverse, often mutually exclusive, temporal dynamics, from smooth trend continuation to nonstationary drift and strict phase-aligned recurrence. While recent deep learning models have improved accuracy, they typically force these diverse patterns through a single computational backbone governed by fixed algorithmic inductive biases (e.g., self-attention or spectral filtering). This single-mechanism approach often struggles with the profound heterogeneity of real-world series, where different variables and forecast horizons necessitate fundamentally different predictive treatments. To address this, we propose GatedLinear: a lightweight framework that frames forecasting as the adaptive routing of complementary linear bases. GatedLinear leverages a pool of three specialized mechanisms: a global trend-seasonal basis for smooth projection, a difference-based incremental basis for nonstationary drift, and a phase-aligned recurrence basis for explicit cyclic reuse. To dynamically orchestrate these distinct behaviors, we introduce a Tri-Factorized Fusion Gate that disentangles routing decisions into channel-specific preferences, horizon-aware offsets, and phase-indexed biases derived from known future time marks. This design allows the model to perform highly granular, point-wise soft routing across different predictive regimes without stacking computationally heavy neural modules. Experiments on standard benchmarks show that our method achieves state-of-the-art or highly competitive accuracy against recent complex foundational models, while offering explicitly interpretable routing patterns and operating with a substantially smaller parameter footprint.

cs.LG↗

arXiv2026

ATOD: Annealed Turn-Aware On-Policy Distillation for Multi-Turn Agentic Tasks

Training small language-model agents for long-horizon interactive tasks requires both fast imitation and reward-driven improvement. On-policy distillation (OPD) provides dense teacher guidance and typically improves rapidly in the early stage, but its gains saturate once the student approaches the teacher, limiting the final performance ceiling. Reinforcement learning (RL) directly optimizes environment rewards and encourages exploratory improvement toward a higher reward-defined ceiling, but sparse and delayed feedback makes early-stage learning much less efficient than OPD. In this paper, we propose ATOD (Annealed Turn-aware On-policy Distillation), a hybrid online distillation algorithm that explicitly exploits this complementarity. (1) ATOD uses an annealed OPD-RL schedule: OPD dominates early training to approach teacher-level behavior, while RL is gradually strengthened to drive reward-based exploration. (2) ATOD introduces Turn-level Disagreement-Uncertainty Reweighting (T-DUR), which softly gates the distillation sig- nal to prioritize turns with high disagreement or uncertainty in long trajectories. Experiments on ALFWorld, WebShop, and Search-QA show that ATOD consistently outperforms competing post-training baselines: across the three student sizes, ATOD improves average success rate by 4.16 points over OPD and 23.62 points over GRPO, while surpassing the corresponding teacher models by 2.16 points.

cs.AI↗

arXiv2026

MBRarefy: data-adaptive multi-bin rarefying for alpha diversity association analysis

Summary: This paper presents MBRarefy, an R package that provides a reproducible workflow for alpha diversity analysis under confounding from heterogeneous library sizes. Building on the multi-bin rarefying approach in Li et al (2024), MBRarefy supports alpha diversity association analysis with repeated rarefying, bin-wise testing, and cross-bin meta-analysis. A key new feature is automated, data-adaptive selection of library size bin thresholds via a genetic algorithm (GA), which replaces ad hoc cutpoints with an objective optimization procedure based on the rarefying-derived profiles. The package also supports routine data-management tasks, including file-based sample-wise processing and standardized output generation, enabling users to execute the full analysis pipeline from raw count files to combined inferential results. Availability and implementation: The R package MBRarefy is freely available on GitHub at https://github.com/mli171/MBRarefy.

stat.CO↗

arXiv2026

Snatcher: Apple Find My Network Exposes Your Lost Devices To Strangers

Apple's Find My network connects nearly one billion devices to locate missing property via Bluetooth Low Energy (BLE). This paper reveals that insecure BLE advertisements and design tradeoffs allow unauthorized discovery and physical theft of lost Apple devices. We develop Snatcher, an attack and analysis framework implemented fully on Android smartphones without specialized hardware. Snatcher identifies vulnerabilities in unencrypted BLE advertisements, unauthenticated acoustic triggers, and slow MAC address randomization. Through three levels - sound-based direction finding, RSSI-IMU sensor-fusion navigation, and spatial-temporal clustering - our Android-based platform physically tracks and locates lost Apple accessories and devices in real-world tests. Our results highlight a crucial conflict between privacy protection, anti-stalking design, and physical security, urging Apple to strengthen Find My defenses.

cs.CR↗

arXiv2026

Changepoint Detection in Categorical Time Series with Application to Daily Total Cloud Cover in Canada

Changepoints are essential for homogenizing categorical time series and analyzing their trends and variations. The original total cloud cover in Canada was recorded hourly in tenths (or eighths), exhibiting inherent seasonality and serial correlation. Lu and Wang (2012) introduced an extended cumulative logit model to detect shifts in the annual frequencies of cloud cover conditions. While annual aggregation mitigates seasonality and serial correlation, it shortens the time series and may lead to overdispersion. This article introduces a marginalized transition model to detect a single changepoint in periodic and serially correlated categorical time series. The model captures serial dependence using a first-order Markov chain and enables category-specific changepoint specification. To enhance computational efficiency, we develop a new parameter estimation procedure for obtaining maximum likelihood estimates. A maximally selected likelihood ratio test statistic is then proposed to test for sudden changes in categorical time series, and the method is illustrated using daily total cloud cover observations recorded at 9 a.m. and 3 p.m. at Fort St. John Airport, British Columbia, Canada.

stat.ME↗

arXiv2026

OpenCompass: A Universal Evaluation Platform for Large Language Models

In recent years, the field of artificial intelligence has undergone a paradigm shift from task-specific small-scale models to general-purpose large language models (LLMs). With the rapid iteration of LLMs, objective, quantitative, and comprehensive evaluation of their capabilities has become a critical link in advancing technological development. Currently, the mainstream static benchmark dataset-based evaluation methods face challenges such as the diversity of task types, inconsistent evaluation criteria, and fragmentation of data and processing workflows, making it difficult to efficiently conduct cross-domain and large-scale model evaluation. To address the aforementioned issues, this paper proposes and open-sources OpenCompass, a one-stop, scalable, and high-concurrency-supported general-purpose LLM evaluation platform. Adhering to the design philosophy of modularization and component decoupling, the platform boasts three core advantages: high compatibility, flexibility, and high concurrency. The core architecture of OpenCompass comprises five key components: the Configuration System, Task Partitioning Module, Execution and Scheduling Module, Task Execution Unit, and Result Visualization Module. Its workflow provides rule-based, LLM-as-a-Judge, and cascaded evaluators to adapt to the requirements of different task scenarios. Supporting mainstream benchmark datasets across multiple domains, including knowledge, reasoning, computation, science, language, code, etc., the platform offers a unified and efficient LLM evaluation tool for both academia and industry, facilitating the accurate identification of strengths and weaknesses of LLMs as well as their subsequent optimization.

cs.CL↗

arXiv2026

2D Optical Beam Scanning using Integrated Acousto-Optics and a Frequency Comb

Optical beam steering is an essential technology for free-space optical communication, reconfigurable optical networks and quantum information systems. Yet conventional steering methods either require bulky mechanical mechanisms, or rely on complex arrays of individually controlled light emitting elements. Integrated acousto-optic beam steering (AOBS) offers non-mechanical, continuous one-dimensional steering on-chip by using traveling acoustic waves with variable frequency to deflect light. In this work, we combine AOBS with an optical frequency comb and optical gratings to enable two-dimensional beam steering from a single aperture. Azimuthal scanning is controlled via acoustic frequency while polar coverage is realized by dispersing frequency comb lines with the gratings. We demonstrate this architecture by sequentially selecting and steering 11 comb lines spanning 1540-1570 nm, achieving a field of view of 18.2 by 4.3 degrees. Validation with a tunable laser extends polar coverage to 11.4 degrees. Both components are realized on the same thin-film lithium niobate platform, providing a pathway toward monolithic integration.

physics.optics↗

arXiv2026

Learning to Commit: Generating Organic Pull Requests via Online Repository Memory

Large language model (LLM)-based coding agents achieve impressive results on controlled benchmarks yet routinely produce pull requests that real maintainers reject. The root cause is not functional incorrectness but a lack of organicity: generated code ignores project-specific conventions, duplicates functionality already provided by internal APIs, and violates implicit architectural constraints accumulated over years of development. Simply exposing an agent to the latest repository snapshot is not enough: the snapshot reveals the final state of the codebase, but not the repository-specific change patterns by which that state was reached. We introduce Learning to Commit, a framework that closes this gap through Online Repository Memory. Given a repository with a strict chronological split, the agent performs supervised contrastive reflection on earlier commits: it blindly attempts to resolve each historical issue, compares its prediction against the oracle diff, and distils the gap into a continuously growing set of skills-reusable patterns capturing coding style, internal API usage, and architectural invariants. When a new PR description arrives, the agent conditions its generation on these accumulated skills, producing changes grounded in the project's own evolution rather than generic pretraining priors. Evaluation is conducted on genuinely future, merged pull requests that could not have been seen during the skill-building phase, and spans multiple dimensions including functional correctness, code-style consistency, internal API reuse rate, and modified-region plausibility. Experiments on an expert-maintained repository with rich commit history show that Online Repository Memory effectively improves organicity scores on held-out future tasks.

cs.SE↗

arXiv2026

Resonance-enhanced integrated acousto-optic beam steering

Optical beam steering is a key technology for free-space optical communication, sensing, and imaging. Mechanical beam steering systems suffer from limited scanning speed and bulky form factors, while existing solid-state solutions rely on pixelated synthetic aperture that requires complex fabrication and control architectures. Integrated acousto-optic beam steering (AOBS) is an emerging technology that enables continuous one-dimensional beam steering using integrated acoustic transducers and fixed-wavelength laser sources. Here, we integrate AOBS with an optical ring resonator on the same thin-film lithium niobate (TFLN) platform to significantly enhance beam steering efficiency and system functionality. The resulting device achieves a resonance-enhanced beam steering efficiency of up to $26\%$ and a field of view of $18^\circ$. Moreover, by leveraging integrated electro-optic control, we dynamically lock the ring-resonator's resonance to a chirped laser frequency, enabling frequency-modulated continuous-wave (FMCW) LiDAR operation. By combining lithium niobate's piezoelectric and electro-optic properties, this work establishes a compact, efficient, and scalable beam-steering platform with co-integrated acousto-optic modulation and electro-optic control for multifunctional applications.

physics.app-ph↗

arXiv2026

Genetic Algorithms in Regression

Many statistical problems involve optimization over a discrete parameter space having an unknown dimension. In such settings, gradient-based methods often fail due to the non-differentiability of the objective function or a non-convex or massive search space with an objective function having many local maxima/minima. This paper presents GAReg, a unified genetic algorithm package that handles discrete optimization regression problems, which works well when standard algorithms are unjustified. GAReg provides a compact chromosome representation supporting optimal knot placement for regression splines, best-subset regression variable selection, and related problems. The package allows for uniform initialization, constraint-preserving crossover and mutation, steady-state replacement, and an optional island-model parallelization. GAReg efficiently searches high-dimensional model spaces, providing near-optimal solutions in settings where exhaustive enumeration or integer or dynamic programming approaches are infeasible.

stat.AP↗

arXiv2026

Kimi K2.5: Visual Agentic Intelligence

We introduce Kimi K2.5, an open-source multimodal agentic model designed to advance general agentic intelligence. K2.5 emphasizes the joint optimization of text and vision so that two modalities enhance each other. This includes a series of techniques such as joint text-vision pre-training, zero-vision SFT, and joint text-vision reinforcement learning. Building on this multimodal foundation, K2.5 introduces Agent Swarm, a self-directed parallel agent orchestration framework that dynamically decomposes complex tasks into heterogeneous sub-problems and executes them concurrently. Extensive evaluations show that Kimi K2.5 achieves state-of-the-art results across various domains including coding, vision, reasoning, and agentic tasks. Agent Swarm also reduces latency by up to $4.5\times$ over single-agent baselines. We release the post-trained Kimi K2.5 model checkpoint to facilitate future research and real-world applications of agentic intelligence.

cs.CL↗

arXiv2026

Temperature-dependent acoustic loss at microwave frequencies in thin-film lithium niobate

Thin-film lithium niobate (TFLN) has emerged as a versatile platform for phononic and photonic devices with applications ranging from classical signal processing to quantum technologies. However, acoustic loss fundamentally limits the performance of acoustic devices on TFLN platforms, yet its physical origin remains insufficiently understood. Here, we systematically investigate acoustic propagation loss in various TFLN platforms, including lithium niobate on insulator (LNOI), lithium niobate on sapphire (LNOS), suspended LN thin films, and bulk LN at gigahertz frequencies over temperatures ranging from 4 K to above room temperature. Using a delay-line method, we extract frequency- and temperature-dependent losses for Rayleigh, shear-horizontal, and Lamb modes. We observe an anomalous non-monotonic temperature dependence in LNOI that closely resembles acoustic loss in amorphous materials, indicating a dominant loss channel associated with the buried oxide layer at low temperatures. At elevated temperatures, the loss converges to the Akhiezer damping governed by phonon-phonon interactions. High-resolution electron microscopy further reveals nanoscale interfacial crystal impurities that may contribute to the increased acoustic loss in TFLN platforms relative to bulk LN. These results elucidate the acoustic loss mechanisms in TFLN and provide guidelines for designing low-loss acoustic devices.

physics.app-ph↗