BestHub
Sign in

Artificial Intelligence

Showing 100 articles max
Latest Popular
Data Party THU
Data Party THU
Oct 12, 2026 · Artificial Intelligence

Beyond Average Accuracy: Representational Equality in Cross-Country LLM Value Simulation

This study evaluates nine open-source LLMs on cross-country value simulation using World Values Survey data across 59 countries, introducing representational equality as a metric to assess whether simulation accuracy is evenly distributed across nations, and tests four intervention strategies for improving both accuracy and equality.

AI fairnessWorld Values Surveybias mitigation
0 likes · 22 min read
Beyond Average Accuracy: Representational Equality in Cross-Country LLM Value Simulation
Data Bricklaying Diary
Data Bricklaying Diary
Oct 12, 2026 · Artificial Intelligence

Agent Observability: Correlating Model, Tools, and Business Receipts to Verify Task Completion

This article argues that Agent observability must go beyond traditional microservice tracing by correlating model calls, retrieval, tool invocations, and control-plane events to task-level identifiers, tracking phases, waits, costs, evidence, and versions, and verifying outcomes with reliable business receipts rather than relying solely on API success metrics.

AI EngineeringAgent observabilityOpenTelemetry
0 likes · 14 min read
Agent Observability: Correlating Model, Tools, and Business Receipts to Verify Task Completion
DataFunTalk
DataFunTalk
Oct 12, 2026 · Artificial Intelligence

Inside Eaton's Equipment Forge: Multi-Agent AI for Complex Manufacturing Decisions

Eaton demonstrates Equipment Forge at Palantir AIPCon 11, a multi-agent system that uses Ontology to connect engineering requirements, BOM, supply chain, and delivery decisions for complex Engineer-to-Order manufacturing, enabling agents to collaboratively analyze trade-offs while humans retain final approval.

AI in manufacturingDigital ThreadEaton
0 likes · 20 min read
Inside Eaton's Equipment Forge: Multi-Agent AI for Complex Manufacturing Decisions
Machine Heart
Machine Heart
Oct 12, 2026 · Artificial Intelligence

RoboFoundry: Self-Evolving System-as-Policy Boosts Embodied Agents Without Model Updates

RoboFoundry freezes foundation models and evolves the surrounding system — context, skills, and execution interfaces — via a closed-loop of execution, reflection, repair, and promotion, achieving large gains on EmbodiedBench, RoboMemArena, LIBERO-PRO, and real-robot zero-shot transfer and online evolution.

Embodied AIEmbodiedBenchLIBERO-PRO
0 likes · 16 min read
RoboFoundry: Self-Evolving System-as-Policy Boosts Embodied Agents Without Model Updates
Machine Heart
Machine Heart
Oct 12, 2026 · Artificial Intelligence

TerraVis: Evaluating World Consistency in Text-to-Image Generation

TerraVis introduces a systematic framework for evaluating world consistency in AI-generated images, covering 18 issue types across object, interaction, and scene levels, validated on 13,500 human annotations across five models.

AI-generated imagesBenchmarkNeurIPS 2026
0 likes · 15 min read
TerraVis: Evaluating World Consistency in Text-to-Image Generation
Data STUDIO
Data STUDIO
Oct 12, 2026 · Artificial Intelligence

Pi Agent Harness: Minimalist AI Coding Agent Returning Control to Developers

This article analyzes Pi, a minimalist terminal-based AI coding agent by Mario Zechner, detailing its harness architecture, six core components, four integration modes, context engineering via AGENTS.md/SYSTEM.md, tree-shaped sessions, extension system, security considerations, and when to choose it over alternatives like Claude Code or Cursor.

AI coding agentAgent Harness EngineeringCodemode
0 likes · 31 min read
Pi Agent Harness: Minimalist AI Coding Agent Returning Control to Developers
Weekly Large Model Application
Weekly Large Model Application
Oct 12, 2026 · Artificial Intelligence

Designing a Telephone-Specific Chinese ASR: Base Model + RL Post-Training Wins

This article presents a system design for a Chinese telephone-specific ASR model, explaining why generic ASR fails on calls due to 8kHz narrowband, codec distortions, and overlapping speech, and proposes a pluggable base model with a four-source data engine, RL post-training using verifiable rewards like entity F1, role-aware decoding, and layered evaluation focused on business-critical metrics rather than raw CER.

ASRChinese ASRRL post-training
0 likes · 12 min read
Designing a Telephone-Specific Chinese ASR: Base Model + RL Post-Training Wins
Linyb Geek Road
Linyb Geek Road
Oct 12, 2026 · Artificial Intelligence

Don't Give Your AI Agent a Super Admin: Palantir Ontology Permission Design

The article explains why enterprise AI agents require fine-grained, runtime permission checks across object types, properties, actions, and identity — not just a simple 'allow AI' switch — using Palantir's ontology framework to illustrate layered security for agents.

AI agentsAction PermissionsObject-Level Security
0 likes · 17 min read
Don't Give Your AI Agent a Super Admin: Palantir Ontology Permission Design
Coder Trainee
Coder Trainee
Oct 11, 2026 · Artificial Intelligence

RAG Evolution: Four Stages from Naive to Agentic (2023-2026) & 2026 Trends

This series finale outlines RAG's evolution through four stages from Naive RAG (2023) to Agentic RAG (2026), identifies key 2026 trends including reasoning-integrated retrieval, multimodal understanding, domain-specific embeddings, and federated knowledge bases, and provides a four-level developer roadmap from beginner to expert.

Agentic RAGGraphRAGRAG
0 likes · 6 min read
RAG Evolution: Four Stages from Naive to Agentic (2023-2026) & 2026 Trends
Machine Heart
Machine Heart
Oct 11, 2026 · Artificial Intelligence

Tsinghua's MGFlow & Elastic Forcing: Distributional Training for One-Step Images & Streaming Video

Tsinghua researchers introduce MGFlow and Elastic Forcing, two distributional training methods that achieve state-of-the-art one-step image generation and enable efficient 14B-parameter streaming video model post-training on just 8 GPUs, using sample distribution matching instead of per-sample supervision.

Elastic ForcingMGFlowTsinghua
0 likes · 20 min read
Tsinghua's MGFlow & Elastic Forcing: Distributional Training for One-Step Images & Streaming Video
Machine Learning Algorithms & Natural Language Processing
Machine Learning Algorithms & Natural Language Processing
Oct 11, 2026 · Artificial Intelligence

DeepSeek's Cost Curve Chronicle: How Token Economics Rewrote LLM Architecture

This article chronicles DeepSeek's paper timeline from 2023 to 2026, analyzing how each model generation optimized training compute, KV cache, and throughput via MoE, MLA, sparse attention, and conditional memory, shifting long-context competition from maximum length to per-token byte cost.

DeepSeekGRPOKV cache optimization
0 likes · 37 min read
DeepSeek's Cost Curve Chronicle: How Token Economics Rewrote LLM Architecture
Architect
Architect
Oct 11, 2026 · Artificial Intelligence

Inside Pi: Minimalist Agent Harness Design & Core Loop Architecture

The article analyzes Pi's agent harness architecture, showing how its core loop stays minimal by handling only model-tool-result cycles while pushing session management, MCP integration, Codemode orchestration, and security boundaries to the edges, enabling extensibility without complicating the main execution path.

AI agentsAgent HarnessCodemode
0 likes · 21 min read
Inside Pi: Minimalist Agent Harness Design & Core Loop Architecture
DataFunSummit
DataFunSummit
Oct 11, 2026 · Artificial Intelligence

Eaton's Equipment Forge: Multi-Agent AI for Complex Engineer-to-Order Decisions

At Palantir AIPCon 11, Eaton demonstrated Equipment Forge, a multi-agent system that uses Palantir's Ontology to coordinate engineering requirements, bill-of-materials checks, supplier lead times, and delivery commitments for custom electrical equipment, keeping humans in the loop for critical decisions while automating cross-domain analysis.

AI in manufacturingDigital ThreadEaton
0 likes · 29 min read
Eaton's Equipment Forge: Multi-Agent AI for Complex Engineer-to-Order Decisions
Data Party THU
Data Party THU
Oct 11, 2026 · Artificial Intelligence

Evolutionary Computation for Trustworthy AI: From Attack-Defense Search to Self-Evolving Systems

This review examines how evolutionary computation supports trustworthy AI across three research lines: evolutionary attacks to find failures, evolutionary defenses to search protective measures, and trustworthy self-evolution to govern which changes persist long-term, highlighting the shift from static robustness to continuous governance of memory, skills, and safeguards.

AI safety benchmarksadversarial attacksco-evolution
0 likes · 31 min read
Evolutionary Computation for Trustworthy AI: From Attack-Defense Search to Self-Evolving Systems
Machine Heart
Machine Heart
Oct 11, 2026 · Artificial Intelligence

USTC's Abductive World Model Gives Robots Causal Reasoning, Runs on Edge at 2-bit

USTC spinout Baize Tongjing develops Abductive World Model (AWM) that predicts then reasons about causes, learns from unlabeled video with 7x data efficiency, and deploys via 2-bit quantization engine BAIZ-RUN on edge chips, validated in three industrial robot scenarios within three months of founding.

2-bit QuantizationAWMAbductive World Model
0 likes · 15 min read
USTC's Abductive World Model Gives Robots Causal Reasoning, Runs on Edge at 2-bit
Machine Heart
Machine Heart
Oct 11, 2026 · Artificial Intelligence

OpenRSI Index: Testing AI Research Agents on Real Tasks with 10K+ GPU Hours

OpenRSI Index evaluates AI agents on real open-source research projects — improving Marin's optimizer, Qwen's RL tasks, and GPIC image generation — using thousands of H100 hours, with isolated workspaces and independent verification to ensure reproducible progress beyond benchmark gaming.

AI research agentsBenchmark EvaluationGPU compute
0 likes · 13 min read
OpenRSI Index: Testing AI Research Agents on Real Tasks with 10K+ GPU Hours
PaperAgent
PaperAgent
Oct 11, 2026 · Artificial Intelligence

Two ACL 2026 Papers Solve Overthinking and Skill Forgetting in LLMs

This article summarizes two high-impact ACL 2026 papers: DRP reduces reasoning token usage by 64% while improving accuracy via skill-aware step pruning, and SAGE doubles skill reuse success rates in agents through sequential rollout and skill-integrated rewards.

ACL 2026DRPLLM reasoning
0 likes · 9 min read
Two ACL 2026 Papers Solve Overthinking and Skill Forgetting in LLMs
SpringMeng
SpringMeng
Oct 11, 2026 · Artificial Intelligence

Agent Developer Interviews: Why Building Demos Isn't Enough for Production

An interviewer shares insights from three Agent developer candidates who had impressive resumes but struggled with production concerns like tool error handling, RAG evaluation, memory design, loop control, and observability, highlighting the gap between demo-building and production-ready systems.

Agent DevelopmentError HandlingInterview Preparation
0 likes · 11 min read
Agent Developer Interviews: Why Building Demos Isn't Enough for Production