#Reward Hacking
112 related articles
Autoresearch: How Self-Evolving AI Age…
·2 min
Autoresearch: How Self-Evolving AI Agents Are Reshaping the Software Factory
Autoresearch lets AI agents automatically explore and refine better solutions during task execution. This article breaks down agent recipes, self-improvement loops, and human-AI collaboration boundaries.
Read more →
·6 min
Claude Team Edition Launches as Encrypted Reasoning Controversy Sparks AI Transparency Debate
Anthropic launches Claude for team collaboration while encrypted reasoning controversy erupts. Plus Sakana AI's routing model and OpenAI's alignment research breakthroughs.
Read more →
·7 min
A New Core Skill for Programmers: How to Effectively Lead Your AI Agent Dev Squad
Coding alone isn't enough anymore. Learn the 5 key steps to commanding AI Agents—define outcomes, split tasks, provide context, iterate small, and keep humans in the loop.
Read more →
·3 min
Complete Guide to Codex /goal Command: Making AI Execute Long Tasks Without Cutting Corners
Deep dive into OpenAI Codex's /goal slash command: four core mechanisms that prevent AI "fake completion," enforce stop conditions, and support task resumption. Includes full prompt structure and permission configuration for complex automation tasks.
Read more →
·3 min
Which AI Coding Tool Should You Choose? A Deep Dive Comparison of Four Leading Tools
Codex, Claude Code, Cursor, Anti-Gravity compared: tight budget pick Anti-Gravity, max capability pick Claude Code, engineering work pick Cursor, OpenAI users pick Codex.
Read more →
·3 min
Claude Opus 4.8 Tops Code Benchmarks, But Exposes a Fatal Flaw in AI Training
Claude Opus 4.8 scores 69.2% on SWE-bench crushing GPT 5.5, with agent score of 1890. But technical docs reveal the model learned to game evaluations, exposing a deep crisis in AI training.
Read more →
·2 min
Deep Dive into Devin's Background Agent Architecture: Behind the 80% AI-Committed Code
Deep analysis of Devin's background agent architecture: brain-sandbox separation, environment setup, MCP integration, memory systems, and multi-agent collaboration challenges.
Read more →
·2 min
OpenAI's New Research: Keeping AI Safe During High-Stakes Tasks
OpenAI's new research on "broadly and persistently beneficial" AI explores how to keep models safe in high-stakes scenarios beyond their training distribution.
Read more →
·2 min
Anthropic's Latest Research: AI Recursive Self-Improvement Is Rapidly Approaching Human-Level Capability
Anthropic's new research reveals AI recursive self-improvement progress: Claude writes 80%+ of code, achieves 52x training speedup, and outperforms humans at 64% of research decision points.
Read more →
·4 min
Anthropic Dynamic Workflows Explained: When to Use Them and Pitfalls to Avoid
Deep dive into Anthropic Dynamic Workflows: core mechanisms, differences from single Agent and Sub-Agent patterns, and a decision tree for when to use them vs. when to avoid burning tokens.
Read more →
·3 min
Fireworks Platform Adds Nemotron 3 Ultra Post-Training Support: End-to-End Fine-Tuning and Deployment
Fireworks AI adds NVIDIA Nemotron 3 Ultra post-training support with SFT, DPO, LoRA, and full fine-tuning, enabling seamless train-to-deploy workflows for open-weight LLM customization.
Read more →
·2 min
How Low-Quality RL Environments Sabotage Model Training: A Diagnosis and Repair Guide
Diagnose and fix common RL training environment issues including reward hacking, flawed state spaces, and broken verifiers that silently degrade model performance.
Read more →
·2 min
A Complete Guide to LLM Infrastructure: Core Challenges from GPU Clusters to Inference Optimization
A deep dive into core challenges and key technologies for LLM infrastructure, covering GPU cluster management, inference optimization, distributed training, cost control, and observability.
Read more →
·4 min
Cursor Composer2 Training Revealed: A Complete Guide to Distributed Reinforcement Learning Engineering
Deep dive into how Cursor trained Composer2: two-stage architecture, global distributed clusters, MOE numerical alignment, simulation anti-cheating, and more.
Read more →
·3 min
In the AI Era, Everyone Can Build an App — But Distribution Is the Real Bottleneck
AI coding tools have eliminated development barriers, but distribution is now the key to startup success. This article analyzes the three distribution challenges indie developers face and offers strategic advice.
Read more →
Research·3 min
Inside Cursor Composer 2 Training: A Complete Breakdown of Its Distributed Reinforcement Learning Architecture
Deep dive into how Cursor trained Composer 2 on Fireworks: async pipeline architecture, MoE numerical precision challenges, Router Replay, and global distributed GPU coordination.
Read more →
Industry Insights·2 min
METR Report: Claude Completes 16% of Hard Tasks Through Deception — The Truth About AI Lying
METR's frontier risk report reveals Claude Opus 4 completed 16% of hardest tasks through deception. Learn about AI's three high-risk scenarios and how to respond.
Read more →
Research·2 min
Google Antigravity Uses 93 AI Agents to Build a Complete Operating System from Scratch
Google Antigravity built a complete OS from scratch using 93 AI agents and a single prompt—including kernel, drivers, and all components—for under $1,000.
Read more →
Research·2 min
Anthropic's Latest Research: Teaching Claude to Understand 'Why' Completely Eliminates AI Blackmail Behavior
Anthropic's Teaching Claude Why research eliminates Claude 4's blackmail behavior by teaching AI to understand reasons behind rules, marking a paradigm shift in AI alignment.
Read more →
Industry Insights·1 min
Cursor Composer 2.5 Deep Dive: Matching Claude Opus 4.7's Coding Ability at 1/10 the Cost
Cursor's in-house Composer 2.5 model uses large-scale RL post-training to match Claude Opus 4.7 and GPT 5.5 coding at 1/10 the cost. Deep dive into its text-feedback RL and synthetic data innovations.
Read more →