Tagged articles
Scaling Law
48 articles · Page 1 of 1
Harvard, NVIDIA, Berkeley, MIT, and Stanford researchers introduce RAX, a framework that scales online exploration for robot agents, achieving 37.4% average improvement over state-of-the-art on five benchmarks without any human demonstrations, using frozen model parameters and self-generated code skills.
Code as PolicyEmbodied AIMulti-Embodiment
0 likes · 15 min read

The TAAC × KDD Cup 2026 workshop in Jeju Island gathered researchers and engineers from academia and leading tech firms to dissect real‑world recommendation challenges, present the competition’s unified sequence‑modeling and feature‑interaction task, and showcase winning solutions that balance effectiveness, efficiency, and scalability.
KDD CupScaling LawTAAC 2026
0 likes · 7 min read

The article presents a 30,000‑hour multimodal tactile‑visual dataset, unified representation models, and two downstream systems (VTLA and TWAM) that demonstrate how large‑scale touch data can dramatically improve robot manipulation success rates and establish tactile perception as a core scaling law for embodied intelligence.
Embodied AIRobot ManipulationScaling Law
0 likes · 9 min read

The article analyzes how agentic architectures, large language models, and generative ranking techniques are applied to overcome high‑concurrency, multimodal, and multi‑hop challenges in modern recommendation and search systems, showcasing concrete designs, performance gains, and real‑world deployments from Alibaba Cloud, Huawei Noah, and Baidu.
AI searchAgentic RAGGenerative Ranking
0 likes · 5 min read

The article presents UserLLM, a three‑stage framework—generative pre‑training, discriminative SFT, and CTR fusion—that leverages ultra‑long user behavior sequences to address the compression and scaling limits of traditional discriminative ranking models, demonstrating significant offline gains and validated scaling‑law behavior across token, layer, and width dimensions.
Scaling LawUserLLMad ranking
0 likes · 17 min read

This week’s tech roundup covers the emergency suspension of OpenAI’s GPT‑5.6, the rapid nine‑month rollout of OpenAI’s Jalapeño AI chip, a Chinese AI‑compute startup’s double‑round financing and KernelCAT platform, shifting AI‑agent paradigms, scaling‑law cautions, new AI design tools, and breakthrough performance of Doubao 2.1 and iFlytek Spark‑X2 in production and education settings.
AI agentsAI chipsDoubao 2.1
0 likes · 32 min read

Galaxy General unveils AstraBrain‑WBC 0.5, a transformer‑based humanoid robot control model that scales from 200 K to 2 billion motion‑capture frames, achieving up to 92.58% tracking success, 0.39 ms latency, and five‑fold speed over TWIST, thereby confirming a scaling law for robot motion control.
AstraBrain-WBCDAgger DistillationHumanoid Robot
0 likes · 16 min read

GigaAI unveiled a dual‑pyramid framework that couples a five‑layer data hierarchy with a three‑layer algorithm hierarchy, demonstrated top‑ranked benchmark results, announced a hundred‑robot home deployment and a 12‑month roadmap toward a physical AGI "GPT‑3 moment".
Dual-Pyramid ArchitectureGigaAIHome Deployment
0 likes · 13 min read

LDA‑1B unifies world modeling and VLA in a latent dynamics action model, ingesting over 30 000 hours of heterogeneous embodied data via a five‑layer AstraData pipeline, employing a unified end‑effector space and quality‑based data allocation, and achieving state‑of‑the‑art success rates on RoboCasa‑GR1 while being fully open‑sourced.
Data IngestionEmbodied AILatent Dynamics
0 likes · 13 min read

Manifold AI’s WorldScape 0.2 achieved the highest overall score on the embodied world‑model benchmark WorldArena, outperforming giants like Google and Nvidia by excelling in comprehensive perception, physics compliance, and 3D accuracy while using only about 10 % of the parameters of competing models, thanks to a newly introduced MoE architecture.
Embodied AIMoEScaling Law
0 likes · 9 min read

A new study by USTC and Shanghai AI Lab uncovers a power‑law scaling relationship between RL fine‑tuning compute and large‑model reasoning performance, offering a quantitative way to predict and control AI capability growth.
AI researchScaling Lawlarge language models
0 likes · 7 min read

The paper presents a systematic empirical study that derives a power‑law scaling formula for reinforcement‑learning‑after‑training of large language models, demonstrating accurate inter‑ and intra‑model performance prediction, learning‑efficiency saturation, data‑reuse benefits, and cross‑architecture validity.
Data ReuseLlama-3Model Efficiency
0 likes · 11 min read

Manifold AI's WorldScape 0.2 topped the WorldArena benchmark by excelling in visual quality, physics compliance, and 3D accuracy, while using only 10% of the parameters of competing models, thanks to a newly introduced MoE architecture that drives a new scaling law for world models.
Embodied AIManifold AIScaling Law
0 likes · 8 min read

The article examines how the dramatic increase in GPU compute power—illustrated by a single H100 GPU equaling about 200 Hadoop instances—challenges the dominance of tree‑based models for structured data, presents scaling‑law experiments with KMLP and FOUND, and argues that pre‑training can redefine the balance between compute, data, and algorithms.
FOUNDGPUKMLP
0 likes · 10 min read

The article examines how a single NVIDIA H100 GPU delivers roughly 200‑fold more FP16 compute than a 96‑core CPU Hadoop node, explores the "Bitter Lesson" of scaling‑driven AI breakthroughs, and presents large‑scale pretraining experiments that show table and sequence models now exhibit clear scaling laws, challenging the dominance of traditional tree‑based approaches.
FOUNDKMLPPretraining
0 likes · 10 min read

This article analyzes DeWu's generative recall system, detailing its background, technical design of the Generative and Rerank models, inference workflow, experimental gains in core consumption and diversity metrics, and future engineering directions such as framework migration, LLM integration, and multimodal generation.
Generative AIRecommendation SystemScaling Law
0 likes · 12 min read

The 2026 KDD Cup, powered by Tencent’s Advertising Algorithm Competition with an $885,000 prize pool, challenges participants to unify sequence modeling and feature interaction in large‑scale recommendation systems, offering academic publication paths, real‑world deployment opportunities, and strict latency constraints that push both research and engineering innovation.
AICompetitionKDD Cup
0 likes · 16 min read

The MIT‑CSAIL paper introduces RandOpt, a single‑step, gradient‑free, fully parallel post‑training algorithm that adds Gaussian noise to pretrained LLM weights and ensembles the results, achieving or surpassing PPO/GRPO performance by exploiting dense "neural thickets" that emerge as model scale grows.
LLMPost-TrainingRandOpt
0 likes · 12 min read

The paper analyzes why enlarging discriminative recommendation models leads to severe over‑fitting, proposes a two‑stage Generative Pretraining for Scalable Discriminative Recommendation (GPSD) that freezes sparse embeddings, and demonstrates across massive e‑commerce data, public benchmarks, and production A/B tests that the approach restores the classic scaling‑law behavior where larger models consistently improve performance.
Scaling LawTransformergenerative pretraining
0 likes · 15 min read

In a deep interview, Baidu AI Computing chief scientist Wang Yanpeng and host Koji trace China's internet infrastructure from the early big‑data era through cloud computing to today's AI boom, highlighting the pivotal role of compute power, GPU acceleration, data scaling, and Baidu's Baige platform in shaping the AI arms race.
AI infrastructureBaidu BaigeGPU computing
0 likes · 26 min read


JD TechNov 6, 2025 · Artificial Intelligence
This article surveys the evolution of generative recommendation systems powered by large language models, detailing their technical foundations, engineering challenges, recent breakthroughs, and future research directions, while highlighting why the paradigm shift is occurring now.
AI EngineeringLLMScaling Law
0 likes · 30 min read

Meta’s recent paper reveals a sigmoid‑shaped scaling law for LLM reinforcement learning, presents extensive 40‑k GPU‑hour experiments, compares various RL designs such as PPO‑off‑policy‑k and Pipeline‑RL‑k, and distills the findings into a practical “ScaleRL” recipe that improves performance and efficiency.
LLMRL OptimizationScaling Law
0 likes · 10 min read


AntTechOct 29, 2025 · Artificial Intelligence
At the Ant Star Innovation Journey event, the Baoling team unveiled their roadmap for trillion‑parameter models, detailing the development of Ling‑1T, Ring‑1T and multimodal Ming series, the scaling‑law‑guided architecture, training innovations, evaluation methods, and open‑source releases that aim to advance efficient, high‑performance AI.
Scaling Lawefficient inferencelarge language model
0 likes · 24 min read

This article explains why Meituan introduced a generative recommendation model, describes the MTGR architecture, data organization, training and inference engines built on TorchRec and TensorRT, reports performance gains and cost reductions, and outlines future directions such as simplifying the recommendation funnel and cross‑business heterogeneous modeling.
Inference OptimizationMTGRScaling Law
0 likes · 15 min read

This article explains Baidu's generative ranking model GRAB, detailing how scaling laws from large language models inspire a new recommendation paradigm, the model's architecture, custom attention mechanisms, training strategies, deployment optimizations, and the resulting business gains in CTR and revenue.
BaiduCTR predictionGenerative AI
0 likes · 22 min read

This article reviews Meta's rotation‑invariant 2‑simplicial attention, explains its trilinear formulation and windowed implementation, analyzes its impact on scaling laws compared with standard dot‑product attention, and presents experimental results showing when the new mechanism offers advantages.
2-simplicial attentionMetaNeural architecture
0 likes · 12 min read

A recent Meta paper introduces a rotation‑invariant 2‑simplicial attention mechanism, demonstrates its superior scaling‑law coefficients over standard dot‑product attention, and provides experimental evidence of improved token efficiency and model performance under constrained token budgets.
2-simplicialAttentionMeta
0 likes · 11 min read


JD TechJun 16, 2025 · Artificial Intelligence
This article showcases three JD tech case studies—using large language models for e‑commerce query expansion, applying sparse large models with scaling‑law experiments to improve ad prediction, and building proactive risk‑prevention systems—to illustrate practical AI engineering that drives higher recall, conversion, and system robustness.
AdvertisingScaling Lawe-commerce
0 likes · 8 min read

Meituan’s recommendation team introduced the MTGR framework, aligning traditional DLRM features with a unified HSTU‑based Transformer to explore scaling laws, delivering a 65‑fold FLOPs boost, 12% lower inference cost, and record gains in online CTR and order volume across its food‑delivery platform.
Inference OptimizationMTGRScaling Law
0 likes · 26 min read

The article analyzes the anti‑scaling phenomenon in video large‑language models, identifies a “temporal hacking” shortcut where models focus on a few key frames, formalizes it via reward‑hacking theory, introduces the Temporal Perplexity (TPL) metric, and proposes an Unhackable Temporal Rewarding (UTR) framework to mitigate the issue.
Reward HackingScaling LawTemporal Perplexity
0 likes · 14 min read

The report analyzes DeepSeek's latest V3 and R1 models, highlights their scaling‑law‑driven cost reductions, explains how Huawei Ascend optimizes inference by cutting KV‑Cache storage and improving compute efficiency, and surveys the model’s deployments across finance, government, manufacturing, and healthcare sectors.
AI EfficiencyAI inferenceDeepSeek
0 likes · 4 min read

In 2024 Alibaba Mama outlines how large‑language models transform search advertising through a three‑line scaling roadmap—explicit inductive‑bias design, implicit compute growth, and auxiliary CV/NLP advances—implemented via a pre‑train/post‑train/CTR paradigm and the LUM user‑behavior model, promising gains in relevance, recall, and real‑time serving while highlighting inference efficiency challenges.
CTR predictionScaling Lawlarge language models
0 likes · 25 min read

Jia Xing’s research introduces Dynamic Margin Selection, a technique that repeatedly refreshes a core set of boundary‑close samples to train large language models efficiently on limited resources, achieving comparable loss to full‑data training, enabling six‑fold model compression, faster inference, and a proposed exponential scaling law for data‑efficient AI.
ICLRLow-Resource TrainingScaling Law
0 likes · 10 min read


ArchitectFeb 19, 2025 · Artificial Intelligence
The article critically examines whether the pre‑training Scaling Law still applies to Grok 3, compares its compute usage and model size with DeepSeek and OpenAI models, evaluates the cost‑effectiveness of pre‑training, RL and test‑time scaling, and explores how these insights shape future large‑language‑model development strategies.
Grok 3Model EfficiencyRL scaling
0 likes · 11 min read


ArchitectFeb 12, 2025 · Artificial Intelligence
The article analyses how S‑shaped growth curves can model the apparent scaling laws of large language models, discusses the three phases of model development, proposes an ability‑density hypothesis, and explores future scenarios where scaling laws may plateau or shift.
AI growthAbility DensityModel Training
0 likes · 16 min read

This article offers a comprehensive overview of large models, covering their historical evolution, technical foundations, the current "hundred‑model" competition, practical use cases across industries, and future challenges such as safety, controllability, and efficient deployment.
Scaling Lawlarge modelsretrieval‑augmented generation
0 likes · 33 min read

The article surveys the AI revolution driven by foundation models and an evolving Scaling Law, outlining four AGI pathways—large models, intelligent robots, brain‑computer interfaces, and digital life—while highlighting transformer‑based convergence, generative‑first‑principle breakthroughs like DeepSeek‑V3, and transformative industry impacts ranging from consumer robots to Medical 2.0, personalized education, and digital‑simulation platforms such as NVIDIA’s Omniverse.
AGIAIAI industry
0 likes · 23 min read


NewBeeNLPDec 3, 2024 · Artificial Intelligence
The article reflects on open‑source LLMs like Qwen2 and Llama 3.1, questioning whether models should self‑review answers, how hidden states might signal uncertainty, the role of loss‑function design, scaling laws, and the trade‑offs between PPO and DPO in alignment.
Scaling LawSelf-Correctionlarge language models
0 likes · 9 min read

This article reviews the evolution of large‑model recommendation techniques, analyzes the specific challenges of health‑oriented e‑commerce recommendation, and details practical deployments such as LLM‑enhanced cold‑start recall, DeepI2I expansion, and scaling‑law‑driven CTR models within JD Health.
CTRRecommendation SystemScaling Law
0 likes · 18 min read

This article provides a comprehensive analysis of the HLLM paper, detailing its hierarchical LLM architecture for item and user modeling, the training objectives, fusion strategies, extensive offline and online experiments, scaling behavior, ablation studies, and practical deployment insights in large‑scale recommendation systems.
LLMScaling LawSequential Modeling
0 likes · 12 min read

This article reviews the evolution of large‑model recommendation techniques, analyzes the specific demands and obstacles of health‑focused e‑commerce, and details JD Health's practical implementations—including LLM‑enhanced recall, deep item‑to‑item models, and scaling‑law‑driven CTR improvements—while discussing open research questions and future directions.
CTRLLM-enhancementScaling Law
0 likes · 17 min read

This paper investigates the embedding dimensional collapse problem that hinders scaling of recommendation models and proposes a Multi-Embedding paradigm that learns multiple embeddings per feature with independent expert networks, demonstrating consistent performance gains across major CTR benchmarks and real‑world ad systems.
CTR predictionScaling Lawartificial intelligence
0 likes · 10 min read

The article details Kuaishou’s development of the 175B “Kuaiyi” multimodal large model, presenting eight novel technical innovations—from Temporal Scaling Law and MiLe Loss to MoE‑enhanced reward modeling—and describes how these advances enable high‑performance AI services such as the AI Xiao Kuai chatbot across diverse real‑world scenarios.
AI applicationsScaling Lawlarge language model
0 likes · 12 min read


NewBeeNLPJul 5, 2024 · Artificial Intelligence
Meta’s new paper introduces the Wukong model, demonstrating that expanding dense‑layer parameters and computational FLOPs in large‑scale recommendation systems follows a clear scaling law, yielding consistent performance gains across massive internal datasets, with detailed analysis of feature modules, parameter impacts, and experimental results.
CTR modelsMetaScaling Law
0 likes · 10 min read

The article analyzes Llama 3’s architecture, training data expansion, model variants, Meta’s open‑source strategy, the evolving gap between open and closed models, and how future breakthroughs in synthetic data will shape scaling laws and large‑model progress through 2025 and beyond.
AI TrendsScaling Lawlarge language models
0 likes · 12 min read


NewBeeNLPApr 10, 2024 · Artificial Intelligence
This article reviews recent scaling‑law research on large‑language‑model fine‑tuning and RLHF, explaining how data quantity, model size, PET parameters, reward‑model size and KL‑penalty affect downstream performance and offering practical insights for efficient training.
LLMRLHFScaling Law
0 likes · 11 min read

Meta introduces a generative recommendation framework (GR) built on the Hierarchical Sequential Transduction Unit (HSTU) that unifies heterogeneous features, treats user behavior as a new modality, and leverages novel encoder and inference optimizations to achieve order‑of‑magnitude scaling in model size, training compute, and online latency while delivering 12‑18% online gains over traditional deep recommendation models.
Generative ModelsHSTUMeta
0 likes · 36 min read

This article presents a memory‑driven architecture (HCNet and MemoNet) that equips recommendation models with scaling‑law characteristics by storing and retrieving arbitrary feature‑combination embeddings, evaluates multi‑hash codebooks, memory‑restoring strategies, key‑feature selection, and demonstrates significant offline and online performance gains.
Scaling Lawfeature interactionlarge language models
0 likes · 15 min read
