Efficient LLM-based Advertising via Model Compression and Parallel Verification

Large language models (LLMs) have shown remarkable potential in advertising scenarios such as ad creative generation and targeted advertising. However, deploying LLMs in real-time advertising systems poses significant challenges due to their high inference latency and computational cost. In this paper, we propose an Efficient Generative Targeting framework that integrates adaptive group quantization, layer-adaptive hierarchical sparsification, and prefix-tree parallel verification to accelerate LLM inference while preserving generation quality. Extensive experiments on two real-world advertising scenarios demonstrate that our framework achieves significant speedup with acceptable quality degradation, making it operationally viable for practical deployments.

Ada-MK: Adaptive MegaKernel Optimization via Automated DAG-based Search for LLM Inference

When large language models (LLMs) serve real-time inference in commercial online advertising systems, end-to-end latency must be strictly bounded to the millisecond range. Yet every token generated during the decode phase triggers thousands of kernel launches, and kernel launch overhead alone can account for 14.6% of end-to-end inference time. MegaKernel eliminates launch overhead and inter-operator HBM round-trips by fusing multiple operators into a single persistent kernel. However, existing MegaKernel implementations face a fundamental tension between portability and efficiency on resource-constrained GPUs such as NVIDIA Ada: hand-tuned solutions are tightly coupled to specific architectures and lack portability, while auto-compiled approaches introduce runtime dynamic scheduling whose branch penalties are unacceptable in latency-critical settings. We observe that under a fixed deployment configuration, the optimal execution path of a MegaKernel is uniquely determined, and runtime dynamic decision-making can be entirely hoisted to compile time. Building on this insight, we propose Ada-MK: (1) a three-dimensional shared-memory constraint model combined with K-dimension splitting that reduces peak shared memory usage by 50%; (2) MLIR-based fine-grained DAG offline search that solidifies the optimal execution path, completely eliminating runtime branching; and (3) a heterogeneous hybrid inference engine that embeds MegaKernel as a plugin into TensorRT-LLM, combining high-throughput Prefill with low-latency Decode. On an NVIDIA L20, Ada-MK improves single-batch throughput by up to 23.6% over vanilla TensorRT-LLM and 50.2% over vLLM, achieving positive gains across all tested scenarios--the first industrial deployment of MegaKernel in a commercial online advertising system.

DVM: A Bytecode Virtual Machine Approach for Dynamic Tensor Computation

Dynamism is common in AI computation, e.g., the dynamic tensor shapes and the dynamic control flows in models. Due to the long compilation time, existing runtime compilation damages the model efficiency, while the offline compilers either suffer from the long compilation time and device memory footprint to cover all the possible execution instances of a dynamic model, or sacrifice optimization opportunities for usability. In this paper, we rethink the feasibility of runtime compilation for dynamic models and identify that the key for it to work is to speed up the compilation or hide the compilation overhead. To do this, we propose a real-time compiler, DVM. In DVM, we design a runtime operator compiler based on a bytecode virtual machine to perform effective and efficient compilation for each dynamic operator instance given its input. Specifically, instead of compiling programs into machine code, we encode the operator program into bytecode on the CPU and decode the bytecode into virtual instructions for direct execution on the NPU. Based on the runtime operator compiler, we further propose an operator fuser, which performs symbol-deduction-based fusion on static graphs and runtime fusion on dynamic graphs. Both pattern- and stacking-based fusion are supported to increase fusion opportunities. Evaluation on operators, subgraphs, and models shows that, compared with TorchInductor, PyTorch-eager and MindSpore-graph-O0, we are up to 11.77$\times$ better in terms of the operator/model efficiency and up to 5 orders of magnitude faster in terms of the maximum compilation time.

DVM: Real-Time Kernel Generation for Dynamic AI Models

Dynamism is common in AI computation, e.g., the dynamic tensor shapes and the dynamic control flows in models. Due to the long compilation time, existing runtime compilation damages the model efficiency, while the offline compilers either suffer from the long compilation time and device memory footprint to cover all the possible execution instances of a dynamic model, or sacrifice optimization opportunities for usability. In this paper, we rethink the feasibility of runtime compilation for dynamic models and identify that the key for it to work is to speed up the compilation or hide the compilation overhead. To do this, we propose a real-time compiler, DVM. In DVM, we design a runtime operator compiler based on a bytecode virtual machine to perform effective and efficient compilation for each dynamic operator instance given its input. Specifically, instead of compiling programs into machine code, we encode the operator program into bytecode on the CPU and decode the bytecode into virtual instructions for direct execution on the NPU. Based on the runtime operator compiler, we further propose an operator fuser, which performs symbol-deduction-based fusion on static graphs and runtime fusion on dynamic graphs. Both pattern- and stacking-based fusion are supported to increase fusion opportunities. Evaluation on operators, subgraphs, and models shows that, compared with TorchInductor, PyTorch-eager and MindSpore-graph-O0, we are up to 11.77$\times$ better in terms of the operator/model efficiency and up to 5 orders of magnitude faster in terms of the maximum compilation time.

From Complex Dynamics to DynFormer: Rethinking Transformers for PDEs

Partial differential equations (PDEs) are fundamental for modeling complex physical systems, yet classical numerical solvers face prohibitive computational costs in high-dimensional and multi-scale regimes. While Transformer-based neural operators have emerged as powerful data-driven alternatives, they conventionally treat all discretized spatial points as uniform, independent tokens. This monolithic approach ignores the intrinsic scale separation of physical fields, applying computationally prohibitive global attention that redundantly mixes smooth large-scale dynamics with high-frequency fluctuations. Rethinking Transformers through the lens of complex dynamics, we propose DynFormer, a novel dynamics-informed neural operator. Rather than applying a uniform attention mechanism across all scales, DynFormer explicitly assigns specialized network modules to distinct physical scales. It leverages a Spectral Embedding to isolate low-frequency modes, enabling a Kronecker-structured attention mechanism to efficiently capture large-scale global interactions with reduced complexity. Concurrently, we introduce a Local-Global-Mixing transformation. This module utilizes nonlinear multiplicative frequency mixing to implicitly reconstruct the small-scale, fast-varying turbulent cascades that are slaved to the macroscopic state, without incurring the cost of global attention. Integrating these modules into a hybrid evolutionary architecture ensures robust long-term temporal stability. Extensive memory-aligned evaluations across four PDE benchmarks demonstrate that DynFormer achieves up to a 95% reduction in relative error compared to state-of-the-art baselines, while significantly reducing GPU memory consumption. Our results establish that embedding first-principles physical dynamics into Transformer architectures yields a highly scalable, theoretically grounded blueprint for PDE surrogate modeling.

TraderBench: How Robust Are AI Agents in Adversarial Capital Markets?

Evaluating AI agents in finance faces two key challenges: static benchmarks require costly expert annotation yet miss the dynamic decision-making central to real-world trading, while LLM-based judges introduce uncontrolled variance on domain-specific tasks. We introduce TraderBench, a benchmark that addresses both issues. It combines expert-verified static tasks (knowledge retrieval, analytical reasoning) with adversarial trading simulations scored purely on realized performance-Sharpe ratio, returns, and drawdown-eliminating judge variance entirely. The framework features two novel tracks: crypto trading with four progressive market-manipulation transforms, and options derivatives scoring across P&L accuracy, Greeks, and risk management. Trading scenarios can be refreshed with new market data to prevent benchmark contamination. Evaluating 13 models (8B open-source to frontier) on ~50 tasks, we find: (1) 8 of 13 models score ~33 on crypto with <1-point variation across adversarial conditions, exposing fixed non-adaptive strategies; (2) extended thinking helps retrieval (+26 points) but has zero impact on trading (+0.3 crypto, -0.1 options). These findings reveal that current agents lack genuine market adaptation, underscoring the need for performance-grounded evaluation in finance.

Seeing the Goal, Missing the Truth: Human Accountability for AI Bias

This research explores how human-defined goals influence the behavior of Large Language Models (LLMs) through purpose-conditioned cognition. Using financial prediction tasks, we show that revealing the downstream use (e.g., predicting stock returns or earnings) of LLM outputs leads the LLM to generate biased sentiment and competition measures, even though these measures are intended to be downstream task-independent. Goal-aware prompting shifts intermediate measures toward the disclosed downstream objective. This purpose leakage improves performance before the LLM's knowledge cutoff, but with no advantage post-cutoff. AI bias due to "seeing the goal" is not an algorithmic flaw, but stems from human accountability in research design to ensure the statistical validity and reliability of AI-generated measurements.

From Classification to Ranking: Enhancing LLM Reasoning Capabilities for MBTI Personality Detection

Personality detection aims to measure an individual's corresponding personality traits through their social media posts. The advancements in Large Language Models (LLMs) offer novel perspectives for personality detection tasks. Existing approaches enhance personality trait analysis by leveraging LLMs to extract semantic information from textual posts as prompts, followed by training classifiers for categorization. However, accurately classifying personality traits remains challenging due to the inherent complexity of human personality and subtle inter-trait distinctions. Moreover, prompt-based methods often exhibit excessive dependency on expert-crafted knowledge without autonomous pattern-learning capacity. To address these limitations, we view personality detection as a ranking task rather than a classification and propose a corresponding reinforcement learning training paradigm. First, we employ supervised fine-tuning (SFT) to establish personality trait ranking capabilities while enforcing standardized output formats, creating a robust initialization. Subsequently, we introduce Group Relative Policy Optimization (GRPO) with a specialized ranking-based reward function. Unlike verification tasks with definitive solutions, personality assessment involves subjective interpretations and blurred boundaries between trait categories. Our reward function explicitly addresses this challenge by training LLMs to learn optimal answer rankings. Comprehensive experiments have demonstrated that our method achieves state-of-the-art performance across multiple personality detection benchmarks.

RIS-Empowered OTFS Modulation With Faster-than-Nyquist Signaling in High-Mobility Wireless Communications

High-mobility wireless communication systems suffer from severe Doppler spread and multi-path delay, which degrade the reliability and spectral efficiency of conventional modulation schemes. Orthogonal time frequency space (OTFS) modulation offers strong robustness in such environments by representing symbols in the delay-Doppler (DD) domain, while faster-than-Nyquist (FTN) signaling can further enhance spectral efficiency through intentional symbol packing. Meanwhile, reconfigurable intelligent surfaces (RIS) provide a promising means to improve link quality via passive beamforming. Motivated by these advantages, we propose a novel RIS-empowered OTFS modulation with FTN signaling (RIS-OTFS-FTN) scheme. First, we establish a unified DD-domain input-output relationship that jointly accounts for RIS passive beamforming, FTN-induced inter-symbol interference, and DD-domain channel characteristics. Based on this model, we provide comprehensive analytical performance for the frame error rate, spectral efficiency, and peak-to-average power ratio (PAPR), etc. Furthermore, a practical RIS phase adjustment strategy with quantized phase selection is designed to maximize the effective channel gain. Extensive Monte Carlo simulations under a standardized extended vehicular A (EVA) channel model validate the theoretical results and provide key insights into the trade-offs among spectral efficiency, PAPR, input back-off (IBO), and error performance, with some interesting insights.The proposed RIS-OTFS-FTN scheme demonstrates notable performance gains in both reliability and spectral efficiency, offering a viable solution for future high-mobility and spectrum-constrained wireless systems.

Negative Entity Suppression for Zero-Shot Captioning with Synthetic Images

Text-only training provides an attractive approach to address data scarcity challenges in zero-shot image captioning (ZIC), avoiding the expense of collecting paired image-text annotations. However, although these approaches perform well within training domains, they suffer from poor cross-domain generalization, often producing hallucinated content when encountering novel visual environments. Retrieval-based methods attempt to mitigate this limitation by leveraging external knowledge, but they can paradoxically exacerbate hallucination when retrieved captions contain entities irrelevant to the inputs. We introduce the concept of negative entities--objects that appear in generated caption but are absent from the input--and propose Negative Entity Suppression (NES) to tackle this challenge. NES seamlessly integrates three stages: (1) it employs synthetic images to ensure consistent image-to-text retrieval across both training and inference; (2) it filters negative entities from retrieved content to enhance accuracy; and (3) it applies attention-level suppression using identified negative entities to further minimize the impact of hallucination-prone features. Evaluation across multiple benchmarks demonstrates that NES maintains competitive in-domain performance while improving cross-domain transfer and reducing hallucination rates, achieving new state-of-the-art results in ZIC. Our code is available at https://github.com/nidongpinyinme/NESCap.