Author

JD Retail Technology

Official platform of JD Retail Technology, delivering insightful R&D news and a deep look into the lives and work of technologists.

820

Articles

Likes

4.4k

Views

Comments

Latest from JD Retail Technology

100 recent articles max

JD Retail Technology

Aug 21, 2026 · Artificial Intelligence

Janus: Dual‑Timescale Scheduling for Production‑Scale Multi‑LLM Serving

Janus, a Service‑Engine co‑design system built on Oxygen xLLM, uses dual‑timescale scheduling, performance‑oracle‑driven placement, and a three‑state model lifecycle to handle bursty traffic, power‑law application hotness, and heterogeneous resource demands, achieving 0.97–1.0 SLO rates while cutting device usage by 27%.

AI InfrastructureLLM servingSOSP 2026

0 likes · 19 min read

Janus: Dual‑Timescale Scheduling for Production‑Scale Multi‑LLM Serving

JD Retail Technology

Aug 10, 2026 · Artificial Intelligence

Bayesian Ensemble (BE): Adaptive Ensemble Selection for Bandits and Reinforcement Learning

The paper introduces Bayesian Ensemble (BE), a lightweight Bayesian layer that dynamically updates the sampling distribution over ensemble members using observed rewards, and extends it to Bayesian Ensemble Bandit (BEB) and Bayesian Ensemble DQN (BE‑DQN), achieving significant regret and click‑through improvements across synthetic, real‑world, and RL benchmarks with minimal computational overhead.

BanditBayesian EnsembleEnsemble Methods

0 likes · 12 min read

Bayesian Ensemble (BE): Adaptive Ensemble Selection for Bandits and Reinforcement Learning

JD Retail Technology

Aug 3, 2026 · Artificial Intelligence

NaviAgent: Tool Orchestration for Oxygen Agents via LLM Planning and Graph Navigation

Large language models struggle with multi‑tool tasks due to lack of global planning; NaviAgent introduces a dual‑layer architecture that separates LLM‑driven interaction decisions from a graph‑based tool navigation model, enabling stable, efficient orchestration of thousands of APIs and achieving higher task success rates across benchmarks.

ICML2026LLM planningdynamic API graph

0 likes · 16 min read

NaviAgent: Tool Orchestration for Oxygen Agents via LLM Planning and Graph Navigation

JD Retail Technology

Jul 27, 2026 · Artificial Intelligence

CVPR 2026: Oxygen XR Team Unveils DirectFisheye‑GS for Native Fisheye Gaussian Splatting

The DirectFisheye‑GS framework embeds a fisheye camera model directly into the 3D Gaussian splatting pipeline and introduces cross‑view joint optimization, eliminating shape distortion and inter‑view inconsistencies, achieving SOTA performance on multiple public datasets while remaining compatible with existing 3DGS toolchains.

3D reconstructionCVPR 2026Cross-View Optimization

0 likes · 10 min read

CVPR 2026: Oxygen XR Team Unveils DirectFisheye‑GS for Native Fisheye Gaussian Splatting

JD Retail Technology

Jul 20, 2026 · Artificial Intelligence

InstEmb: Instruction-Following Embeddings that Anticipate Future Model Outputs

InstEmb introduces learnable look‑ahead tokens and dual‑anchor alignment pooling to inject output‑aware semantics into instruction‑following embeddings without extra decoding, achieving strong gains on ICML‑2026 benchmarks such as FollowIR and InfoSearch while maintaining competitive performance on general sentence‑embedding tasks.

EmbeddingICML 2026LLaMA-3

0 likes · 15 min read

InstEmb: Instruction-Following Embeddings that Anticipate Future Model Outputs

JD Retail Technology

Jul 13, 2026 · Artificial Intelligence

Inside JD’s Oxygen AIIC: An Industrial‑Scale LLM/VLM‑Powered Product Knowledge Platform for Billions of SKUs

JD’s Oxygen AIIC combines human‑in‑the‑loop ontology engineering, a semantic search‑then‑discrimination pipeline, and a self‑evolving multi‑task LLM/VLM model to produce high‑quality product knowledge for over a hundred thousand categories and billions of daily SKU updates, boosting search coverage to 80%, attribute auto‑fill to over 80%, cutting quality issues by 37% and raising click‑through by 9% while achieving 94.2% precision and 82.8% recall.

JD.comKnowledge GraphLLM

0 likes · 21 min read

Inside JD’s Oxygen AIIC: An Industrial‑Scale LLM/VLM‑Powered Product Knowledge Platform for Billions of SKUs

JD Retail Technology

Jul 6, 2026 · Artificial Intelligence

GROLE: Instance-Level Expert Routing for Incremental Learning in Oxygen AIIC Models

The paper introduces GROLE, a two‑stage incremental‑learning framework that builds a frozen pool of task‑specific LoRA experts and trains a lightweight instance‑level selector via reinforcement‑learning‑based gradient‑free optimization with Dirichlet sampling, achieving superior stability‑plasticity trade‑offs and state‑of‑the‑art results on multiple CL benchmarks.

GROLELoRAincremental learning

0 likes · 14 min read

GROLE: Instance-Level Expert Routing for Incremental Learning in Oxygen AIIC Models

JD Retail Technology

Jun 29, 2026 · Artificial Intelligence

Uni-AdGen: Unified Autoregressive Model for Personalized Image‑Text Ad Generation (CVPR 2026)

Uni‑AdGen unifies image and text generation in a single autoregressive framework, introduces a coarse‑to‑fine preference module and foreground‑aware control, and demonstrates superior performance on the million‑scale PAd1M dataset with novel evaluation metrics for personalized advertising.

Autoregressive Modelevaluation metricslarge-scale dataset

0 likes · 15 min read

Uni-AdGen: Unified Autoregressive Model for Personalized Image‑Text Ad Generation (CVPR 2026)

JD Retail Technology

Jun 25, 2026 · Artificial Intelligence

JD Donates Oxygen xLLM Inference Engine to OpenAtom Foundation to Accelerate Domestic AI Infra

JD announced the donation of its self‑developed Oxygen xLLM large‑model inference engine to the OpenAtom Open Source Foundation, detailing its service‑engine decoupled architecture, performance breakthroughs, multi‑chip support, and early industrial validations that aim to foster a collaborative domestic AI infrastructure ecosystem.

AI inferenceDomestic AI ecosystemEngineering Intelligence

0 likes · 8 min read

JD Donates Oxygen xLLM Inference Engine to OpenAtom Foundation to Accelerate Domestic AI Infra

JD Retail Technology

Jun 23, 2026 · Artificial Intelligence

How RAD‑DPO Aligns Preferences in OxygenSearch Generative Retrieval (SIGIR 2026)

This article analyzes the challenges of generative retrieval for e‑commerce (shared SID prefixes, noisy implicit feedback, and probability squeezing) and presents RAD‑DPO, a robust adaptive denoising direct preference optimization that uses session‑level multi‑label contrast, token‑level gradient detachment, and dynamic reward weighting to improve both effectiveness and training efficiency.

RAD-DPOe-commerce searchgenerative retrieval

0 likes · 18 min read

How RAD‑DPO Aligns Preferences in OxygenSearch Generative Retrieval (SIGIR 2026)