arXiv subjects

Yiren Zhao

Publications and source records attributed to Yiren Zhao.

At least 37 records · Page 2Linked to original sources

arXiv2026

Team of Thoughts: Efficient Test-time Scaling of Agentic Systems through Orchestrated Tool Calling

Existing Multi-Agent Systems (MAS) typically rely on homogeneous model configurations, failing to exploit the diverse expertise inherent in different post-trained architectures. We propose Team-of-Thoughts, a heterogeneous MAS framework that treats diverse models as specialized tools within an orchestrator-driven paradigm. Team-of-Thoughts introduces two novel components: (1) Orchestrator Calibration, which identifies models with superior coordination and synthesis capabilities, and (2) Agent Self-Assessment, a protocol where tool agents profile their own domain-specific strengths to guide selection. At inference, the orchestrator dynamically activates the most compatible agents based on these profiles to maximize capability coverage. Across five mathematical reasoning and code generation benchmarks, Team-of-Thoughts consistently outperforms individual models and existing MAS baselines. Notably, on AIME24 and LiveCodeBench, Team-of-Thoughts achieves 96.00% and 77.91% accuracy, respectively, significantly improving over homogeneous role-play baselines (80.00% and 65.93%).

cs.CL↗

arXiv2026

On the Existence and Behavior of Secondary Attention Sinks

Attention sinks are tokens, often the beginning-of-sequence (BOS) token, that receive disproportionately high attention despite limited semantic relevance. In this work, we identify a class of attention sinks, which we term secondary sinks, that differ fundamentally from the sinks studied in prior works, which we term primary sinks. While prior works have identified that tokens other than BOS can sometimes become sinks, they were found to exhibit properties analogous to the BOS token. Specifically, they emerge at the same layer, persist throughout the network and draw a large amount of attention mass. Whereas, we find the existence of secondary sinks that arise primarily in middle layers and can persist for a variable number of layers, and draw a smaller, but still significant, amount of attention mass. Through extensive experiments across 11 model families, we analyze where these secondary sinks appear, their properties, how they are formed, and their impact on the attention mechanism. Specifically, we show that: (1) these sinks are formed by specific middle-layer MLP modules; these MLPs map token representations to vectors that align with the direction of the primary sink of that layer. (2) The $\ell_2$-norm of these vectors determines the sink score of the secondary sink, and also the number of layers it lasts for, thereby leading to different impacts on the attention mechanisms accordingly. (3) The primary sink weakens in middle layers, coinciding with the emergence of secondary sinks. We observe that in larger-scale models, the location and lifetime of the sinks, together referred to as sink levels, appear in a more deterministic and frequent manner. Specifically, we identify three sink levels in QwQ-32B and six levels in Qwen3-14B. We open-sourced our findings at github.com/JeffreyWong20/Secondary-Attention-Sinks.

cs.LG↗

arXiv2026

Q-StaR: A Quasi-Static Routing Scheme for NoCs

In networks-on-chip, static routing schemes are favored for their simplicity and predictability, but they cannot effectively balance network load due to the unawareness of runtime load distribution. Q-StaR discovers two factors (topology and traffic distribution) that determine the long-term trend of load distribution, and proposes N-Rank to extract this trend. The obtained information is used to guide BiDOR's route selection at runtime, thereby improving load balancing while retaining simplicity and predictability. Simulation validates that Q-StaR significantly outperforms the typical dimension-order routing (throughput under uniform traffic improved by 42.9\%, and mean/maximum latency under realistic workloads reduced by 86.4\%/95.3\%).

cs.NI↗

arXiv2026

Deep Kernel Fusion for Transformers

Agentic LLM inference with long contexts is increasingly limited by memory bandwidth rather than compute. In this setting, SwiGLU MLP blocks, whose large weights exceed cache capacity, become a major yet under-optimized bottleneck. We propose DeepFusionKernel, a deeply fused kernel that cuts HBM traffic and boosts cache reuse, delivering up to 13.2% speedup on H100 and 9.7% on A100 over SGLang. Integrated with SGLang and paired with a kernel scheduler, DeepFusionKernel ensures consistent accelerations over generation lengths, while remaining adaptable to diverse models, inference configurations, and hardware platforms.

cs.LG↗

arXiv2026

Heterogeneous Computing: The Key to Powering the Future of AI Agent Inference

AI agent inference is driving an inference heavy datacenter future and exposes bottlenecks beyond compute - especially memory capacity, memory bandwidth and high-speed interconnect. We introduce two metrics - Operational Intensity (OI) and Capacity Footprint (CF) - that jointly explain regimes the classic roofline analysis misses, including the memory capacity wall. Across agentic workflows (chat, coding, web use, computer use) and base model choices (GQA/MLA, MoE, quantization), OI/CF can shift dramatically, with long context KV cache making decode highly memory bound. These observations motivate disaggregated serving and system level heterogeneity: specialized prefill and decode accelerators, broader scale up networking, and decoupled compute-memory enabled by optical I/O. We further hypothesize agent-hardware co design, multiple inference accelerators within one system, and high bandwidth, large capacity memory disaggregation as foundations for adaptation to evolving OI/CF. Together, these directions chart a path to sustain efficiency and capability for large scale agentic AI inference.

cs.AI↗

arXiv2025

Reasoning Introduces New Poisoning Attacks Yet Makes Them More Complicated

Early research into data poisoning attacks against Large Language Models (LLMs) demonstrated the ease with which backdoors could be injected. More recent LLMs add step-by-step reasoning, expanding the attack surface to include the intermediate chain-of-thought (CoT) and its inherent trait of decomposing problems into subproblems. Using these vectors for more stealthy poisoning, we introduce ``decomposed reasoning poison'', in which the attacker modifies only the reasoning path, leaving prompts and final answers clean, and splits the trigger across multiple, individually harmless components. Fascinatingly, while it remains possible to inject these decomposed poisons, reliably activating them to change final answers (rather than just the CoT) is surprisingly difficult. This difficulty arises because the models can often recover from backdoors that are activated within their thought processes. Ultimately, it appears that an emergent form of backdoor robustness is originating from the reasoning capabilities of these advanced LLMs, as well as from the architectural separation between reasoning and final answer generation.

cs.CR↗

arXiv2025

Scaling Laws For Mixed Quantization

Post-training quantization of Large Language Models (LLMs) has proven effective in reducing the memory and computational requirements for inference. In this study, we focus on a straightforward question: When aiming for a target accuracy or perplexity with low-precision quantization, how much high-precision computation needs to be preserved, and how fine-grained this quantization would need to be as we scale LLMs to larger sizes? We first introduce two critical metrics, named the quantization ratio ($Q_r$) and quantization block size ($Q_b$). The former measures the number of parameters quantized to low-precision arithmetic normalized by the total parameter count, whereas the latter defines the number of values within a block that share a scaling factor, akin to the block size concept introduced in the FP4 format in NVIDIA's Blackwell architecture. Through extensive and carefully controlled experiments across different models and quantization methods, we propose a unified scaling law on post-training quantization (PTQ) that can predict loss degeneration for varying $Q_r$ and $Q_b$. For $Q_r$, our scaling law implies that parameter scaling and ratio scaling have a multiplicative relationship. Consequently, larger models are more amenable to a higher quantization ratio $Q_r$, thus supporting an increase in the adoption of mixed quantization for inference. Regarding $Q_b$, our findings indicate that a small block size, similar to that used in Blackwell, is not essential for large models. Employing a small $Q_b$ can instead unnecessarily complicate the design of the hardware circuit.

cs.CL↗

arXiv2025

Hardware and Software Platform Inference

It is now a common business practice to buy access to large language model (LLM) inference rather than self-host, because of significant upfront hardware infrastructure and energy costs. However, as a buyer, there is no mechanism to verify the authenticity of the advertised service including the serving hardware platform, e.g. that it is actually being served using an NVIDIA H100. Furthermore, there are reports suggesting that model providers may deliver models that differ slightly from the advertised ones, often to make them run on less expensive hardware. That way, a client pays premium for a capable model access on more expensive hardware, yet ends up being served by a (potentially less capable) cheaper model on cheaper hardware. In this paper we introduce hardware and software platform inference (HSPI) -- a method for identifying the underlying GPU architecture and software stack of a (black-box) machine learning model solely based on its input-output behavior. Our method leverages the inherent differences of various GPU architectures and compilers to distinguish between different GPU types and software stacks. By analyzing the numerical patterns in the model's outputs, we propose a classification framework capable of accurately identifying the GPU used for model inference as well as the underlying software configuration. Our findings demonstrate the feasibility of inferring GPU type from black-box models. We evaluate HSPI against models served on different real hardware and find that in a white-box setting we can distinguish between different GPUs with between $83.9\%$ and $100\%$ accuracy. Even in a black-box setting we achieve results that are up to 3x higher than random guess accuracy. Our code is available at https://github.com/ChengZhang-98/HSPI.

cs.LG↗

arXiv2025

Refining Salience-Aware Sparse Fine-Tuning Strategies for Language Models

Parameter-Efficient Fine-Tuning (PEFT) has gained prominence through low-rank adaptation methods like LoRA. In this paper, we focus on sparsity-based PEFT (SPEFT), which introduces trainable sparse adaptations to the weight matrices in the model, offering greater flexibility in selecting fine-tuned parameters compared to low-rank methods. We conduct the first systematic evaluation of salience metrics for SPEFT, inspired by zero-cost NAS proxies, and identify simple gradient-based metrics is reliable, and results are on par with the best alternatives, offering both computational efficiency and robust performance. Additionally, we compare static and dynamic masking strategies, finding that static masking, which predetermines non-zero entries before training, delivers efficiency without sacrificing performance, while dynamic masking offers no substantial benefits. Across NLP tasks, a simple gradient-based, static SPEFT consistently outperforms other fine-tuning methods for LLMs, providing a simple yet effective baseline for SPEFT. Our work challenges the notion that complexity is necessary for effective PEFT, while our open-source framework establishes a reproducible benchmark for future research, which is available at [https://github.com/0-ml/speft].

cs.CL↗

arXiv2025

Mixture of Weight-shared Heterogeneous Group Attention Experts for Dynamic Token-wise KV Optimization

Transformer models face scalability challenges in causal language modeling (CLM) due to inefficient memory allocation for growing key-value (KV) caches, which strains compute and storage resources. Existing methods like Grouped Query Attention (GQA) and token-level KV optimization improve efficiency but rely on rigid resource allocation, often discarding "low-priority" tokens or statically grouping them, failing to address the dynamic spectrum of token importance. We propose mixSGA, a novel mixture-of-expert (MoE) approach that dynamically optimizes token-wise computation and memory allocation. Unlike prior approaches, mixSGA retains all tokens while adaptively routing them to specialized experts with varying KV group sizes, balancing granularity and efficiency. Our key novelties include: (1) a token-wise expert-choice routing mechanism guided by learned importance scores, enabling proportional resource allocation without token discard; (2) weight-sharing across grouped attention projections to minimize parameter overhead; and (3) an auxiliary loss to ensure one-hot routing decisions for training-inference consistency in CLMs. Extensive evaluations across Llama3, TinyLlama, OPT, and Gemma2 model families show mixSGA's superiority over static baselines. On instruction-following and continued pretraining tasks, mixSGA achieves higher ROUGE-L and lower perplexity under the same KV budgets.

cs.CL↗

arXiv2025

LLM4DV: Using Large Language Models for Hardware Test Stimuli Generation

Hardware design verification (DV) is a process that checks the functional equivalence of a hardware design against its specifications, improving hardware reliability and robustness. A key task in the DV process is the test stimuli generation, which creates a set of conditions or inputs for testing. These test conditions are often complex and specific to the given hardware design, requiring substantial human engineering effort to optimize. We seek a solution of automated and efficient testing for arbitrary hardware designs that takes advantage of large language models (LLMs). LLMs have already shown promising results for improving hardware design automation, but remain under-explored for hardware DV. In this paper, we propose an open-source benchmarking framework named LLM4DV that efficiently orchestrates LLMs for automated hardware test stimuli generation. Our analysis evaluates six different LLMs involving six prompting improvements over eight hardware designs and provides insight for future work on LLMs development for efficient automated DV.

cs.LG↗

arXiv2025

Locking Machine Learning Models into Hardware

Modern machine learning (ML) models are expensive IP and business competitiveness often depends on keeping this IP confidential. This in turn restricts how these models are deployed; for example, it is unclear how to deploy a model on-device without inevitably leaking the underlying model. At the same time, confidential computing technologies such as multi-party computation or homomorphic encryption remain impractical for wide adoption. In this paper, we take a different approach and investigate the feasibility of ML-specific mechanisms that deter unauthorized model use by restricting the model to only be usable on specific hardware, making adoption on unauthorized hardware inconvenient. That way, even if IP is compromised, it cannot be trivially used without specialised hardware or major model adjustment. In a sense, we seek to enable cheap \emph{locking of machine learning models into specific hardware}. We demonstrate that \emph{locking} mechanisms are feasible by either targeting efficiency of model representations, making such models incompatible with quantization, or tying the model's operation to specific characteristics of hardware, such as the number of clock cycles for arithmetic operations. We demonstrate that locking comes with negligible overheads, while significantly restricting usability of the resultant model on unauthorized hardware.

cs.CR↗

arXiv2025

AMPLE: Event-Driven Accelerator for Mixed-Precision Inference of Graph Neural Networks

Graph Neural Networks (GNNs) have recently gained attention due to their performance on non-Euclidean data. The use of custom hardware architectures proves particularly beneficial for GNNs due to their irregular memory access patterns, resulting from the sparse structure of graphs. However, existing FPGA accelerators are limited by their double buffering mechanism, which doesn't account for the irregular node distribution in typical graph datasets. To address this, we introduce \textbf{AMPLE} (Accelerated Message Passing Logic Engine), an FPGA accelerator leveraging a new event-driven programming flow. We develop a mixed-arithmetic architecture, enabling GNN inference to be quantized at a node-level granularity. Finally, prefetcher for data and instructions is implemented to optimize off-chip memory access and maximize node parallelism. Evaluation on citation and social media graph datasets ranging from $2$K to $700$K nodes showed a mean speedup of $243\times$ and $7.2\times$ against CPU and GPU counterparts, respectively.

cs.AR↗

arXiv2025

ARIES: Autonomous Reasoning with LLMs on Interactive Thought Graph Environments

Recent research has shown that LLM performance on reasoning tasks can be enhanced by scaling test-time compute. One promising approach, particularly with decomposable problems, involves arranging intermediate solutions as a graph on which transformations are performed to explore the solution space. However, prior works rely on pre-determined, task-specific transformation schedules which are subject to a set of searched hyperparameters. In this work, we view thought graph transformations as actions in a Markov decision process, and implement policy agents to drive effective action policies for the underlying reasoning LLM agent. In particular, we investigate the ability for another LLM to act as a policy agent on thought graph environments and introduce ARIES, a multi-agent architecture for reasoning with LLMs. In ARIES, reasoning LLM agents solve decomposed subproblems, while policy LLM agents maintain visibility of the thought graph states, and dynamically adapt the problem-solving strategy. Through extensive experiments, we observe that using off-the-shelf LLMs as policy agents with no supervised fine-tuning (SFT) can yield up to $29\%$ higher accuracy on HumanEval relative to static transformation schedules, as well as reducing inference costs by $35\%$ and avoid any search requirements. We also conduct a thorough analysis of observed failure modes, highlighting that limitations on LLM sizes and the depth of problem decomposition can be seen as challenges to scaling LLM-guided reasoning.

cs.AI↗

arXiv2025

QERA: an Analytical Framework for Quantization Error Reconstruction

The growing number of parameters and computational demands of large language models (LLMs) present significant challenges for their efficient deployment. Recently, there is an increasing interest in quantizing weights to extremely low precision while offsetting the resulting error with low-rank, high-precision error reconstruction terms. The combination of quantization and low-rank approximation is now popular in both adapter-based, parameter-efficient fine-tuning methods such as LoftQ and low-precision inference techniques including ZeroQuant-V2. Usually, the low-rank terms are calculated via the singular value decomposition (SVD) of the weight quantization error, minimizing the Frobenius and spectral norms of the weight approximation error. Recent methods like LQ-LoRA and LQER introduced hand-crafted heuristics to minimize errors in layer outputs (activations) rather than weights, resulting improved quantization results. However, these heuristic methods lack an analytical solution to guide the design of quantization error reconstruction terms. In this paper, we revisit this problem and formulate an analytical framework, named Quantization Error Reconstruction Analysis (QERA), and offer a closed-form solution to the problem. We show QERA benefits both existing low-precision fine-tuning and inference methods -- QERA achieves a fine-tuned accuracy gain of $Δ_{\text{acc}}$ = 6.05% of 2-bit RoBERTa-base on GLUE compared to LoftQ; and obtains $Δ_{\text{acc}}$ = 2.97% higher post-training quantization accuracy of 4-bit Llama-3.1-70B on average than ZeroQuant-V2 and $Δ_{\text{ppl}}$ = - 0.28 lower perplexity on WikiText2 than LQER.

cs.LG↗

arXiv2025

Cached Multi-Lora Composition for Multi-Concept Image Generation

Low-Rank Adaptation (LoRA) has emerged as a widely adopted technique in text-to-image models, enabling precise rendering of multiple distinct elements, such as characters and styles, in multi-concept image generation. However, current approaches face significant challenges when composing these LoRAs for multi-concept image generation, resulting in diminished generated image quality. In this paper, we initially investigate the role of LoRAs in the denoising process through the lens of the Fourier frequency domain. Based on the hypothesis that applying multiple LoRAs could lead to "semantic conflicts", we find that certain LoRAs amplify high-frequency features such as edges and textures, whereas others mainly focus on low-frequency elements, including the overall structure and smooth color gradients. Building on these insights, we devise a frequency domain based sequencing strategy to determine the optimal order in which LoRAs should be integrated during inference. This strategy offers a methodical and generalizable solution compared to the naive integration commonly found in existing LoRA fusion techniques. To fully leverage our proposed LoRA order sequence determination method in multi-LoRA composition tasks, we introduce a novel, training-free framework, Cached Multi-LoRA (CMLoRA), designed to efficiently integrate multiple LoRAs while maintaining cohesive image generation. With its flexible backbone for multi-LoRA fusion and a non-uniform caching strategy tailored to individual LoRAs, CMLoRA has the potential to reduce semantic conflicts in LoRA composition and improve computational efficiency. Our experimental evaluations demonstrate that CMLoRA outperforms state-of-the-art training-free LoRA fusion methods by a significant margin -- it achieves an average improvement of $2.19\%$ in CLIPScore, and $11.25\%$ in MLLM win rate compared to LoraHub, LoRA Composite, and LoRA Switch.

cs.CV↗

arXiv2025

Omni-DNA: A Unified Genomic Foundation Model for Cross-Modal and Multi-Task Learning

Large Language Models (LLMs) demonstrate remarkable generalizability across diverse tasks, yet genomic foundation models (GFMs) still require separate finetuning for each downstream application, creating significant overhead as model sizes grow. Moreover, existing GFMs are constrained by rigid output formats, limiting their applicability to various genomic tasks. In this work, we revisit the transformer-based auto-regressive models and introduce Omni-DNA, a family of cross-modal multi-task models ranging from 20 million to 1 billion parameters. Our approach consists of two stages: (i) pretraining on DNA sequences with next token prediction objective, and (ii) expanding the multi-modal task-specific tokens and finetuning for multiple downstream tasks simultaneously. When evaluated on the Nucleotide Transformer and GB benchmarks, Omni-DNA achieves state-of-the-art performance on 18 out of 26 tasks. Through multi-task finetuning, Omni-DNA addresses 10 acetylation and methylation tasks at once, surpassing models trained on each task individually. Finally, we design two complex genomic tasks, DNA2Function and Needle-in-DNA, which map DNA sequences to textual functional descriptions and images, respectively, indicating Omni-DNA's cross-modal capabilities to broaden the scope of genomic applications. All the models are available through https://huggingface.co/collections/zehui127

q-bio.GN↗

arXiv2024

GV-Rep: A Large-Scale Dataset for Genetic Variant Representation Learning

Genetic variants (GVs) are defined as differences in the DNA sequences among individuals and play a crucial role in diagnosing and treating genetic diseases. The rapid decrease in next generation sequencing cost has led to an exponential increase in patient-level GV data. This growth poses a challenge for clinicians who must efficiently prioritize patient-specific GVs and integrate them with existing genomic databases to inform patient management. To addressing the interpretation of GVs, genomic foundation models (GFMs) have emerged. However, these models lack standardized performance assessments, leading to considerable variability in model evaluations. This poses the question: How effectively do deep learning methods classify unknown GVs and align them with clinically-verified GVs? We argue that representation learning, which transforms raw data into meaningful feature spaces, is an effective approach for addressing both indexing and classification challenges. We introduce a large-scale Genetic Variant dataset, named GV-Rep, featuring variable-length contexts and detailed annotations, designed for deep learning models to learn GV representations across various traits, diseases, tissue types, and experimental contexts. Our contributions are three-fold: (i) Construction of a comprehensive dataset with 7 million records, each labeled with characteristics of the corresponding variants, alongside additional data from 17,548 gene knockout tests across 1,107 cell types, 1,808 variant combinations, and 156 unique clinically verified GVs from real-world patients. (ii) Analysis of the structure and properties of the dataset. (iii) Experimentation of the dataset with pre-trained GFMs. The results show a significant gap between GFMs current capabilities and accurate GV representation. We hope this dataset will help advance genomic deep learning to bridge this gap.

cs.LG↗