arXiv subjects

Jianzhong Guo

Publications and source records attributed to Jianzhong Guo.

8 recordsLinked to original sources

arXiv2026

DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning

General reasoning represents a long-standing and formidable challenge in artificial intelligence. Recent breakthroughs, exemplified by large language models (LLMs) and chain-of-thought prompting, have achieved considerable success on foundational reasoning tasks. However, this success is heavily contingent upon extensive human-annotated demonstrations, and models' capabilities are still insufficient for more complex problems. Here we show that the reasoning abilities of LLMs can be incentivized through pure reinforcement learning (RL), obviating the need for human-labeled reasoning trajectories. The proposed RL framework facilitates the emergent development of advanced reasoning patterns, such as self-reflection, verification, and dynamic strategy adaptation. Consequently, the trained model achieves superior performance on verifiable tasks such as mathematics, coding competitions, and STEM fields, surpassing its counterparts trained via conventional supervised learning on human demonstrations. Moreover, the emergent reasoning patterns exhibited by these large-scale models can be systematically harnessed to guide and enhance the reasoning capabilities of smaller models.

cs.CL↗

arXiv2025

DeepSeek-V3 Technical Report

We present DeepSeek-V3, a strong Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token. To achieve efficient inference and cost-effective training, DeepSeek-V3 adopts Multi-head Latent Attention (MLA) and DeepSeekMoE architectures, which were thoroughly validated in DeepSeek-V2. Furthermore, DeepSeek-V3 pioneers an auxiliary-loss-free strategy for load balancing and sets a multi-token prediction training objective for stronger performance. We pre-train DeepSeek-V3 on 14.8 trillion diverse and high-quality tokens, followed by Supervised Fine-Tuning and Reinforcement Learning stages to fully harness its capabilities. Comprehensive evaluations reveal that DeepSeek-V3 outperforms other open-source models and achieves performance comparable to leading closed-source models. Despite its excellent performance, DeepSeek-V3 requires only 2.788M H800 GPU hours for its full training. In addition, its training process is remarkably stable. Throughout the entire training process, we did not experience any irrecoverable loss spikes or perform any rollbacks. The model checkpoints are available at https://github.com/deepseek-ai/DeepSeek-V3.

cs.CL↗

arXiv2024

Fire-Flyer AI-HPC: A Cost-Effective Software-Hardware Co-Design for Deep Learning

The rapid progress in Deep Learning (DL) and Large Language Models (LLMs) has exponentially increased demands of computational power and bandwidth. This, combined with the high costs of faster computing chips and interconnects, has significantly inflated High Performance Computing (HPC) construction costs. To address these challenges, we introduce the Fire-Flyer AI-HPC architecture, a synergistic hardware-software co-design framework and its best practices. For DL training, we deployed the Fire-Flyer 2 with 10,000 PCIe A100 GPUs, achieved performance approximating the DGX-A100 while reducing costs by half and energy consumption by 40%. We specifically engineered HFReduce to accelerate allreduce communication and implemented numerous measures to keep our Computation-Storage Integrated Network congestion-free. Through our software stack, including HaiScale, 3FS, and HAI-Platform, we achieved substantial scalability by overlapping computation and communication. Our system-oriented experience from DL training provides valuable insights to drive future advancements in AI-HPC.

cs.DC↗

arXiv2024

DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model

We present DeepSeek-V2, a strong Mixture-of-Experts (MoE) language model characterized by economical training and efficient inference. It comprises 236B total parameters, of which 21B are activated for each token, and supports a context length of 128K tokens. DeepSeek-V2 adopts innovative architectures including Multi-head Latent Attention (MLA) and DeepSeekMoE. MLA guarantees efficient inference through significantly compressing the Key-Value (KV) cache into a latent vector, while DeepSeekMoE enables training strong models at an economical cost through sparse computation. Compared with DeepSeek 67B, DeepSeek-V2 achieves significantly stronger performance, and meanwhile saves 42.5% of training costs, reduces the KV cache by 93.3%, and boosts the maximum generation throughput to 5.76 times. We pretrain DeepSeek-V2 on a high-quality and multi-source corpus consisting of 8.1T tokens, and further perform Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) to fully unlock its potential. Evaluation results show that, even with only 21B activated parameters, DeepSeek-V2 and its chat versions still achieve top-tier performance among open-source models.

cs.CL↗

arXiv2024

DeepSeek LLM: Scaling Open-Source Language Models with Longtermism

The rapid development of open-source large language models (LLMs) has been truly remarkable. However, the scaling law described in previous literature presents varying conclusions, which casts a dark cloud over scaling LLMs. We delve into the study of scaling laws and present our distinctive findings that facilitate scaling of large scale models in two commonly used open-source configurations, 7B and 67B. Guided by the scaling laws, we introduce DeepSeek LLM, a project dedicated to advancing open-source language models with a long-term perspective. To support the pre-training phase, we have developed a dataset that currently consists of 2 trillion tokens and is continuously expanding. We further conduct supervised fine-tuning (SFT) and Direct Preference Optimization (DPO) on DeepSeek LLM Base models, resulting in the creation of DeepSeek Chat models. Our evaluation results demonstrate that DeepSeek LLM 67B surpasses LLaMA-2 70B on various benchmarks, particularly in the domains of code, mathematics, and reasoning. Furthermore, open-ended evaluations reveal that DeepSeek LLM 67B Chat exhibits superior performance compared to GPT-3.5.

cs.CL↗

arXiv2023

Multi-level Relation Learning for Cross-domain Few-shot Hyperspectral Image Classification

Cross-domain few-shot hyperspectral image classification focuses on learning prior knowledge from a large number of labeled samples from source domains and then transferring the knowledge to the tasks which contain few labeled samples in target domains. Following the metric-based manner, many current methods first extract the features of the query and support samples, and then directly predict the classes of query samples according to their distance to the support samples or prototypes. The relations between samples have not been fully explored and utilized. Different from current works, this paper proposes to learn sample relations on different levels and take them into the model learning process, to improve the cross-domain few-shot hyperspectral image classification. Building on current method of "Deep Cross-Domain Few-Shot Learning for Hyperspectral Image Classification" which adopts a domain discriminator to deal with domain-level distribution difference, the proposed method applies contrastive learning to learn the class-level sample relations to obtain more discriminable sample features. In addition, it adopts a transformer based cross-attention learning module to learn the set-level sample relations and acquire the attention from query samples to support samples. Our experimental results have demonstrated the contribution of the multi-level relation learning mechanism for few-shot hyperspectral image classification when compared with the state of the art methods.

cs.CV↗

arXiv2023

Sonogenetics is a novel antiarrhythmic treatment

Arrhythmia of the heart is a dangerous and potentially fatal condition. The current widely used treatment is the implantable cardioverter defibrillator (ICD), but it is invasive and affects the patient's quality of life. The sonogenetic treatment technique proposed here focuses transthoracic ultrasound on the heart, noninvasively stimulates endogenous stretch-activated Piezo1 ion channels on the focal region's cardiomyocyte plasma membrane, and restores normal heart rhythm. In contrast to anchoring the implanted ICD lead at a fixed position in the myocardium, the size and position of the ultrasound focal region can be selected dynamically by adjusting the signal phases of every piezoelectric chip on the wearable ultrasonic phased array, and it allows novel and efficient defibrillations. Based on the developed interdisciplinary electromechanical model of sonogenetic treatment, our analysis shows that the proposed ultrasound intensity and frequency will be safe and painless for humans and well below the limits established by the U.S. Food and Drug Administration.

physics.med-ph↗

arXiv2023

Reconstructing bifurcation diagrams of chaotic circuits with reservoir computing

Model-free reconstruction of the bifurcation diagrams of Chua's circuits by the technique of parameter-aware reservoir computing is investigated. We demonstrate that: (1) reservoir computer can be utilized as a noise filter to recover the system trajectory from noisy signals; (2) for a single Chua circuit, the machine trained by the noisy time series measured at several sampling states is capable of reconstructing the whole bifurcation diagram of the circuit with a high precision; (3) for two coupled chaotic Chua circuits of mismatched parameters, the machine trained by the noisy time series measured at several coupling strengths is able to anticipate the variation of the synchronization degree of the coupled circuits with respect to the coupling strength over a wide range. The studies verify the capability of the technique of parameter-aware reservoir computing in learning the dynamics of chaotic circuits from noisy signals, signifying the potential application of this technique in reconstructing the bifurcation diagram of real-world chaotic systems.

cond-mat.dis-nn↗