Search arXiv⌕ Search

arXiv subjects

Yi Liu

Publications and source records attributed to Yi Liu.

At least 181 records · Page 10Linked to original sources
arXiv2026

OVAL: Open-Vocabulary Augmented Memory Model for Lifelong Object Goal Navigation

Object Goal Navigation (ObjectNav) refers to an agent navigating to an object in an unseen environment, which is an ability often required in the accomplishment of complex tasks. While existing methods demonstrate proficiency in isolated single object navigation, their limitations emerge in the restricted applicability of lifelong memory representations, which ultimately hinders effective navigation toward continual targets over extended periods. To address this problem, we propose OVAL, a novel lifelong open-vocabulary memory framework, which enables efficient and precise execution of long-term navigation in semantically open tasks. Within this framework, we introduce memory descriptors to facilitate structured management of the memory model. Additionally, we propose a novel probability-based exploration strategy, utilizing a multi-value frontier scoring to enhance lifelong exploration efficiency. Extensive experiments demonstrate the efficiency and robustness of the proposed system.

cs.RO↗
arXiv2026

BankerToolBench: Evaluating AI Agents in End-to-End Investment Banking Workflows

Existing AI benchmarks lack the fidelity to assess economically meaningful progress on professional workflows. To evaluate frontier AI agents in a high-value, labor-intensive profession, we introduce BankerToolBench (BTB): an open-source benchmark of end-to-end analytical workflows routinely performed by junior investment bankers. To develop an ecologically valid benchmark grounded in representative work environments, we collaborated with 502 investment bankers from leading firms. BTB requires agents to execute senior banker requests by navigating data rooms, using industry tools (market data platform, SEC filings database), and generating multi-file deliverables--including Excel financial models, PowerPoint pitch decks, and PDF/Word reports. Completing a BTB task takes bankers up to 21 hours, underscoring the economic stakes of successfully delegating this work to AI. BTB enables automated evaluation of any LLM or agent, scoring deliverables against 100+ rubric criteria defined by veteran investment bankers to capture stakeholder utility. Testing 9 frontier models, we find that even the best-performing model (GPT-5.4) fails nearly half of the rubric criteria and bankers rate 0% of its outputs as client-ready. Our failure analysis reveals key obstacles (such as breakdowns in cross-artifact consistency) and improvement directions for agentic AI in high-stakes professional workflows.

cs.AI↗
arXiv2026

Optimal Adaptive Market Making: A Theoretical Framework for High-Yield Liquidity Provision in Perpetual Futures Markets

We develop a rigorous theoretical framework for optimal market making in perpetual futures markets with zero maker fees. We model the market maker's problem as a stochastic optimal control problem on a filtered probability space, where the controls are adaptive bid-ask spreads and inventory hedging decisions across two exchanges. Our contributions include: (i) a PnL decomposition theorem separating revenue into spread income, adverse selection loss, inventory carrying cost, hedging friction, and funding rate exposure; (ii) the Hamilton-Jacobi-Bellman equation for the joint spread-inventory-hedging control problem under CARA utility with a verification theorem; (iii) High-APY Regime Theorems characterizing profitable regions via five dimensionless parameters, culminating in a Master APY Formula; (iv) analysis of zero-fee economics on decentralized perpetual exchanges with optimal entry-exit thresholds; (v) optimal cross-exchange hedging policies with funding rate dynamics and a hedge regime trichotomy; (vi) a robustness margin quantifying parameter uncertainty tolerance; (vii) exponential drawdown probability bounds and a universal APY-VaR identity; (viii) ergodic inventory distribution under optimal control with Bayesian adaptive estimation; (ix) Kelly-optimal leverage with ruin boundaries; and (x) multi-pair portfolio allocation with diversification saturation results. Numerical analysis with twenty-three figures reveals phase transitions between profitable and unprofitable regimes. Our framework unifies and extends the Avellaneda-Stoikov, Gueant-Lehalle-Fernandez-Tapia, and Glosten-Milgrom paradigms for modern decentralized venue microstructure.

cs.AI↗
arXiv2026

PaddleOCR-VL-1.5: Towards a Multi-Task 0.9B VLM for Robust In-the-Wild Document Parsing

We introduce PaddleOCR-VL-1.5, an upgraded model achieving a new state-of-the-art (SOTA) accuracy of 94.5% on OmniDocBench v1.5. To rigorously evaluate robustness against real-world physical distortions, including scanning, skew, warping, screen-photography, and illumination, we propose the Real5-OmniDocBench benchmark. Experimental results demonstrate that this enhanced model attains SOTA performance on the newly curated benchmark. Furthermore, we extend the model's capabilities by incorporating seal recognition and text spotting tasks, while remaining a 0.9B ultra-compact VLM with high efficiency. Code: https://github.com/PaddlePaddle/PaddleOCR

cs.CV↗
arXiv2026

Boosting Document Parsing Efficiency and Performance with Coarse-to-Fine Visual Processing

Document parsing is a fine-grained task where image resolution significantly impacts performance. While advanced research leveraging vision-language models benefits from high-resolution input to boost model performance, this often leads to a quadratic increase in the number of vision tokens and significantly raises computational costs. We attribute this inefficiency to substantial visual regions redundancy in document images, like background. To tackle this, we propose PaddleOCR-VL, a novel coarse-to-fine architecture that focuses on semantically relevant regions while suppressing redundant ones, thereby improving both efficiency and performance. Specifically, we introduce a lightweight Valid Region Focus Module (VRFM) which leverages localization and contextual relationship prediction capabilities to identify valid vision tokens. Subsequently, we design and train a compact yet powerful 0.9B vision-language model (PaddleOCR-VL-0.9B) to perform detailed recognition, guided by VRFM outputs to avoid direct processing of the entire large image. Extensive experiments demonstrate that PaddleOCR-VL achieves state-of-the-art performance in both page-level parsing and element-level recognition. It significantly outperforms existing solutions, exhibits strong competitiveness against top-tier VLMs, and delivers fast inference while utilizing substantially fewer vision tokens and parameters, highlighting the effectiveness of targeted coarse-to-fine parsing for accurate and efficient document understanding. The source code and models are publicly available at https://github.com/PaddlePaddle/PaddleOCR.

cs.CV↗
arXiv2026

NavCrafter: Exploring 3D Scenes from a Single Image

Creating flexible 3D scenes from a single image is vital when direct 3D data acquisition is costly or impractical. We introduce NavCrafter, a novel framework that explores 3D scenes from a single image by synthesizing novel-view video sequences with camera controllability and temporal-spatial consistency. NavCrafter leverages video diffusion models to capture rich 3D priors and adopts a geometry-aware expansion strategy to progressively extend scene coverage. To enable controllable multi-view synthesis, we introduce a multi-stage camera control mechanism that conditions diffusion models with diverse trajectories via dual-branch camera injection and attention modulation. We further propose a collision-aware camera trajectory planner and an enhanced 3D Gaussian Splatting (3DGS) pipeline with depth-aligned supervision, structural regularization and refinement. Extensive experiments demonstrate that NavCrafter achieves state-of-the-art novel-view synthesis under large viewpoint shifts and substantially improves 3D reconstruction fidelity.

cs.CV↗
arXiv2026

Synonymous Codon Usage Bias Overrides Phylogeny to Reflect Convergent Frond Architecture in a Rapidly Radiating Fern Family Thelypteridaceae

Convergent evolution provides powerful evidence for natural selection, yet its molecular basis is typically sought in protein-coding amino acid substitutions. Whether adaptive pressures can drive the convergent evolution of synonymous codon usage bias (CUB) to override phylogenetic history remains a fundamental question. Here, we investigate this within the rapidly radiating fern family Thelypteridaceae by establishing a comparative framework that integrates chloroplast phylogenomics with dimensionality reduction of codon usage, morphological data, and divergence time estimation. Our results reveal that chloroplast CUB patterns are strikingly incongruent with the phylogeny of this family. Instead, they partition species into distinct clusters that strongly correlate with a convergently evolved morphological trait, lamina base architecture, a key adaptation whose radiation we date to the early Neogene. This convergent molecular signal is driven by a specific subset of photosynthesis-related genes (ndhJ, psaA, and psbD), which exhibit a high density of type-specific, third-position codon substitutions. These findings demonstrate that CUB can serve as a powerful, quantifiable indicator of adaptive history, revealing a cryptic layer of molecular convergence linked to the regulation of protein synthesis. Our work providing a new framework for uncovering adaptive histories obscured by complex evolutionary processes.

q-bio.PE↗
arXiv2026

Supply-Chain Poisoning Attacks Against LLM Coding Agent Skill Ecosystems

LLM-based coding agents extend their capabilities via third-party agent skills distributed through open marketplaces without mandatory security review. Unlike traditional packages, these skills are executed as operational directives with system-level privileges, so a single malicious skill can compromise the host. Prior work has not examined whether supply-chain attacks can directly hijack an agent's action space, such as file writes, shell commands, and network requests, despite existing safeguards. We introduce Document-Driven Implicit Payload Execution (DDIPE), which embeds malicious logic in code examples and configuration templates within skill documentation. Because agents reuse these examples during normal tasks, the payload executes without explicit prompts. Using an LLM-driven pipeline, we generate 1,070 adversarial skills from 81 seeds across 15 MITRE ATTACK categories. Across four frameworks and five models, DDIPE achieves 11.6% to 33.5% bypass rates, while explicit instruction attacks achieve 0% under strong defenses. Static analysis detects most cases, but 2.5% evade both detection and alignment. Responsible disclosure led to four confirmed vulnerabilities and two fixes.

cs.CR↗
arXiv2026

The Spectral Geometry of Thought: Phase Transitions, Instruction Reversal, Token-Level Dynamics, and Perfect Correctness Prediction in How Transformers Reason

We discover that large language models exhibit \emph{spectral phase transitions} in their hidden activation spaces when engaging in reasoning versus factual recall. Through systematic spectral analysis across \textbf{11 models} spanning \textbf{5 architecture families} (Qwen, Pythia, Phi, Llama, DeepSeek-R1), we identify \textbf{seven} core phenomena: (1)~\textbf{Reasoning Spectral Compression} -- 9/11 models show significantly lower $α$ for reasoning ($p < 0.05$), with larger effects in stronger models; (2)~\textbf{Instruction Tuning Spectral Reversal} -- base models show reasoning $α< $ factual $α$, while instruction-tuned models reverse this relationship; (3)~\textbf{Architecture-Dependent Generation Taxonomy} -- prompt-to-response shifts partition into expansion, compression, and equilibrium regimes; (4)~\textbf{Spectral Scaling Law} -- $α_\text{reasoning} \propto -0.074 \ln N$ across 4 Qwen base models ($R^2 = 0.46$); (5)~\textbf{Token-Level Spectral Cascade} -- per-token alpha tracking reveals local synchronization that decays exponentially with layer distance, and is weaker for reasoning than factual tasks; (6)~\textbf{Reasoning Step Spectral Punctuation} -- phase-transition signatures align with reasoning step boundaries; and (7)~\textbf{Spectral Correctness Prediction} -- spectral $α$ alone achieves AUC $= 1.000$ (Qwen2.5-7B, late layers) and mean AUC $= 0.893$ across 6 models in predicting correctness \emph{before} the final answer is generated. Together, these findings establish a comprehensive \emph{spectral theory of reasoning} in transformers, revealing that the geometry of thought is universal in direction, architecture-specific in dynamics, and predictive of outcome.

cs.LG↗
arXiv2026

The Spectral Lifecycle of Transformer Training: Transient Compression Waves, Persistent Spectral Gradients, and the Q/K--V Asymmetry

We present the first systematic study of weight matrix singular value spectra \emph{during} transformer pretraining, tracking full SVD decompositions of every weight matrix at 25-step intervals across three model scales (30M--285M parameters). We discover three phenomena: \textbf{(1)~Transient Compression Waves:} stable rank compression propagates as a traveling wave from early to late layers, creating a dramatic gradient that peaks early then \emph{reverses} -- late layers eventually over-compress past early layers. \textbf{(2)~Persistent Spectral Gradients:} the power-law exponent~$α$ develops a permanent depth gradient forming a non-monotonic inverted-U in deeper models, with peaks shifting toward earlier layers as depth increases. \textbf{(3)~Q/K--V Functional Asymmetry:} value/output projections compress uniformly while query/key projections carry the full depth-dependent dynamics. The dissociation between transient compression and persistent spectral shape reveals that \emph{rank and spectral shape encode fundamentally different information about training}. We formalize this as a two-timescale dynamical model and derive scaling laws ($Δα\propto L^{0.26}$, $R^2{=}0.99$). We validate on nine models across three families (custom, GPT-2, Pythia; 30M--1B parameters; 8--36 layers), demonstrate that $α$ predicts layer importance ($ρ{=}0.69$--$0.84$, $p{<}0.02$), and show that spectral-guided pruning outperforms Last-N heuristics by $1.1{\times}$--$3.6{\times}$ across seven models in two families (GPT-2 124M--774M, Pythia 160M--1B), with worst-vs-best gaps up to $23.7{\times}$ confirming the causal role of spectral structure.

cs.LG↗
arXiv2026

Altermagnetism and Room-Temperature Metal-to-Insulator Transition in CsCr$_2$S$_2$O

Metal-to-insulator transitions (MITs), particularly near room temperature, have been extensively studied in nonmagnetic and conventional ferromagnetic and antiferromagnetic systems, yet the co-emergence of MIT and altermagnetism (AM) remains unexplored. Here, a layered chromium-based compound CsCr$_2$S$_2$O that realizes this coexistence was synthesized. It crystalizes in CeCr$_2$Si$_2$C-type structure with Cr moments orders in a C-type antiferromagnetic configuration below $T_\mathrm{N}$ = 326 K, constituting a room-temperature d-wave altermagnet. In the altermagnetic state, a subsequent Verwey-type MIT appears at $T_\mathrm{MI}$ = 305 K, driven by a tetragonal-to-orthorhombic structural distortion and stripe charge ordering of Cr$^{+2}$/Cr$^{+3}$ ions, while maintaining its altermagnetic character. First-principles calculations show moment-dependent spin-split electronic structures with maximum splitting energies of ~0.6 eV and ~0.3 eV in the metallic and insulating states, respectively. Our work links the two prominent phenomena, MIT and AM, in a single material, establishing a new platform for potential spintronic applications.

cond-mat.mtrl-sci↗
arXiv2026

A Conditional Generative Framework for Synthetic Data Augmentation in Segmenting Thin and Elongated Structures in Biological Images

Thin and elongated filamentous structures, such as microtubules and actin filaments, often play important roles in biological systems. Segmenting these filaments in biological images is a fundamental step for quantitative analysis. Recent advances in deep learning have significantly improved the performance of filament segmentation. However, there is a big challenge in acquiring high quality pixel-level annotated dataset for filamentous structures, as the dense distribution and geometric properties of filaments making manual annotation extremely laborious and time-consuming. To address the data shortage problem, we propose a conditional generative framework based on the Pix2Pix architecture to generate realistic filaments in microscopy images from binary masks. We also propose a filament-aware structural loss to improve the structure similarity when generating synthetic images. Our experiments have demonstrated the effectiveness of our approach and outperformed existing model trained without synthetic data.

cs.CV↗
arXiv2026

ACoT-VLA: Action Chain-of-Thought for Vision-Language-Action Models

Vision-Language-Action models have emerged as essential generalist robot policies for diverse manipulation tasks, conventionally relying on directly translating multimodal inputs into actions via Vision-Language Model embeddings. Recent advancements have introduced explicit intermediary reasoning-such as sub-task prediction (language) or goal image synthesis (vision)-to guide action generation. However, these intermediate reasoning are often indirect and inherently limited in their capacity to convey the full, granular information required for precise action execution. Instead, we posit that the most effective form of reasoning is one that deliberates directly in the action space. We introduce Action Chain-of-Thought (ACoT), a paradigm where the reasoning process itself is formulated as a structured sequence of coarse action intents that guide the final policy. In this paper, we propose ACoT-VLA, a novel architecture that materializes the ACoT paradigm. Specifically, we introduce two complementary components: an Explicit Action Reasoner (EAR) and Implicit Action Reasoner (IAR). The former proposes coarse reference trajectories as explicit action-level reasoning steps, while the latter extracts latent action priors from internal representations of multimodal input, co-forming an ACoT that conditions the downstream action head to enable grounded policy learning. Extensive experiments in real-world and simulation environments demonstrate the superiority of our proposed method. Code is available at: https://github.com/AgibotTech/ACoT-VLA.

cs.RO↗
arXiv2026

Tomonaga-Luttinger liquid and charge-density wave in a quasi-one-dimensional material

In one-dimensional (1D) electron systems, the Fermi liquid state breaks down due either to electron interactions, which results in a Tomonaga-Luttinger liquid (TLL) state, or to Peierls instability, which leads to an insulating charge-density-wave (CDW) phase. In general, these two phenomena are mutually exclusive, and their coexistence remains elusive in real materials. Here, we report the discovery of a new quasi-1D material, Cs$_{1-δ}$Cr$_3$S$_3$, which unexpectedly exhibits coexistence of the antithetical CDW and TLL states. The CDW state is evidenced by the intra-unit-cell dimerization, and the opening of an optical band gap of $\sim$250 meV. Meanwhile, TLL behaviour is unambiguously demonstrated by the measurements of electrical transport and angle-resolved photoemission spectroscopy, which reveal a power-law scaling with temperature, bias voltage and electron energy. Band structure calculations reveal isolated, linearly dispersive, 1D bands around the Fermi level. For the dimerized CDW phase, the 1D Fermi-surface sheets located at the boundary of the Brillouin zone are gapped from intra-unit-cell bond symmetry breaking. Experimentally, subtle Cs vacancies shift the Fermi level into the linearly dispersive valence band, enabling the observation of TLL behaviour without interrupting the CDW order. This work establishes Cs$_{1-δ}$Cr$_3$S$_3$ as a rare material platform in which the antagonistic Fermi-liquid instabilities coexist and intertwine, opening new avenues for studying emergent quantum phenomena in 1D systems.

cond-mat.str-el↗
arXiv2026

Cs$_3$V$_9$Te$_{13}$: A Correlated Electron System with Topological Flat Bands

Correlated electron systems with topological flat bands show great promise in exploring exotic quantum phenomena. However, such crystalline materials remain rare. Here we report the discovery of a novel material, Cs$_3$V$_9$Te$_{13}$, which unexpectedly exhibits magnetism and significant electron correlations. The crystal structure features two interpenetrating sets of vanadium triangles that can be linked with an ideal kagome lattice. The physical property measurements demonstrate a cascade of correlated electron phenomena, including quasi-two-dimensional bad metal, non-Fermi-liquid behavior, antiferromagnetic spin-density-wave transition at $T_\mathrm{N}$ = 47 K, possible short-range spin ordering at $\sim$350 K, a large Sommerfeld coefficient of 246 mJ mol-fu$^{-1}$ K$^{-2}$, and pressure-induced quantum criticality. These correlated electron behaviors are associated with the topological flat bands at the Fermi level, the latter of which are generated from the V2 sublattice in terms of a bipartite kagome model. Our findings establish Cs$_3$V$_9$Te$_{13}$ as a brand new correlated matter that synergistically combines flat-band physics and tunable properties.

cond-mat.str-el↗
arXiv2026

Time is Not Compute: Scaling Laws for Wall-Clock Constrained Training on Consumer GPUs

Scaling laws relate model quality to compute budget (FLOPs), but practitioners face wall-clock time constraints, not compute budgets. We study optimal model sizing under fixed time budgets from 5 minutes to 24 hours on consumer GPUs (RTX 4090). Across 70+ runs spanning 50M--1031M parameters, we find: (1)~at each time budget a U-shaped curve emerges where too-small models overfit and too-large models undertrain; (2)~optimal model size follows $N^* \propto t^{0.60}$, growing \emph{faster} than Chinchilla's $N^* \propto C^{0.50}$, with $α= 0.60 \pm 0.07$ robustly exceeding compute-optimal across all sensitivity analyses; (3)~a \emph{dual U-shape mechanism}: short-budget U-curves arise from compute bottlenecks, while long-budget U-curves emerge from data bottlenecks (overfitting), with an intermediate regime where the U-curve temporarily disappears. These findings have immediate implications for researchers training on consumer hardware, where wall-clock time -- not FLOPs -- is the binding constraint. We release all code, logs, and 70+ experimental configurations.

cs.PF↗
arXiv2026

Principles of Conditionality and Layering of Error Rates with Application to Platform Trials

There has been a misconception that only one type of error rate control is necessary in clinical trials, leading to debates over whether to prioritize Familywise Error Rate (FWER) or False Discovery Rate (FDR). This misconception has led to misleading statements about FWER control and proposals to shift towards FDR control, which could be manipulated by the industry. In reality, since the early 2000s, biopharmaceutical statistics have implicitly applied two layers of Type I error rate control. This aligns with Tukey's 1953 invention of Error Rate per Family (ERpF) for controlling error across studies, while FWER applies within each study. Our paper clarifies this layering, using Platform trials to demonstrate the verifiable conditions needed across studies for the FDA to fulfill its regulatory mission. We show that controlling FWER within a study at $5\%$ inherently controls ERpF across studies at 5-per-100, regardless of study correlations. This supports current regulatory practices that protect public health while fostering innovation. We also address concerns about ERpF stability in Platform trials, where shared controls introduce dependencies. By applying the Conditionality Principle and utilizing an innovative Shiny app, we explore how correlations impact ERpF variability, providing deeper insights for informed decision-making. Our findings, supported by principles like Layering of Error Rate Controls and the Conditionality Principle, are particularly relevant as Platform trials gain popularity for their efficiency in testing multiple treatments simultaneously.

stat.ME↗
arXiv2026

PP-OCRv5: A Specialized 5M-Parameter Model Rivaling Billion-Parameter Vision-Language Models on OCR Tasks

The advent of "OCR 2.0" and large-scale vision-language models (VLMs) has set new benchmarks in text recognition. However, these unified architectures often come with significant computational demands, challenges in precise text localization within complex layouts, and a propensity for textual hallucinations. Revisiting the prevailing notion that model scale is the sole path to high accuracy, this paper introduces PP-OCRv5, a meticulously optimized, lightweight OCR system with merely 5 million parameters. We demonstrate that PP-OCRv5 achieves performance competitive with many billion-parameter VLMs on standard OCR benchmarks, while offering superior localization precision and reduced hallucinations. The cornerstone of our success lies not in architectural expansion but in a data-centric investigation. We systematically dissect the role of training data by quantifying three critical dimensions: data difficulty, data accuracy, and data diversity. Our extensive experiments reveal that with a sufficient volume of high-quality, accurately labeled, and diverse data, the performance ceiling for traditional, efficient two-stage OCR pipelines is far higher than commonly assumed. This work provides compelling evidence for the viability of lightweight, specialized models in the large-model era and offers practical insights into data curation for OCR. The source code and models are publicly available at https://github.com/PaddlePaddle/PaddleOCR.

cs.CV↗