https://knowledged.to/tags/ai/Recent content in Ai on knowledged.toHugoen-usMon, 25 May 2026 23:31:16 +0530https://knowledged.to/ai/concepts/anti-narration/Mon, 25 May 2026 18:01:07 +0000https://knowledged.to/ai/concepts/anti-narration/Harness pattern that forces verification before accepting fluent AI outputs as correct.https://knowledged.to/notes/ml/ppo-proximal-policy-optimization/Tue, 19 May 2026 17:18:44 +0000https://knowledged.to/notes/ml/ppo-proximal-policy-optimization/Overview of PPO, the clipped policy-gradient RL algorithm used in RLHF for InstructGPT and original ChatGPT.https://knowledged.to/notes/ml/grpo-group-relative-policy-optimization/Tue, 19 May 2026 17:17:58 +0000https://knowledged.to/notes/ml/grpo-group-relative-policy-optimization/Critic-free RL algorithm that replaces PPO's value model with group-relative rewards for LLM fine-tuning.https://knowledged.to/notes/ml/tool-dc-strategic-anchor-grouping-example/Tue, 19 May 2026 06:12:48 +0000https://knowledged.to/notes/ml/tool-dc-strategic-anchor-grouping-example/Concrete web-search example showing how Tool-DC strategic anchor grouping reduces schema-confusion in tool calls.https://knowledged.to/notes/ml/agentflow/Tue, 19 May 2026 05:08:59 +0000https://knowledged.to/notes/ml/agentflow/Overview of AgentFlow, an agent architecture that trains a planner with Flow-GRPO for multi-turn tool use.https://knowledged.to/notes/ml/tool-dc-framework/Tue, 19 May 2026 03:53:48 +0000https://knowledged.to/notes/ml/tool-dc-framework/Overview of Tool-DC, a try-check-retry framework for robust long-context tool-calling with large tool registries.https://knowledged.to/notes/ml/attention/Sun, 17 May 2026 05:54:45 +0000https://knowledged.to/notes/ml/attention/Explanation of the attention mechanism in ML, covering Query/Key/Value, self-attention, multi-head, causal, cross-attention, and efficiency variants like FlashAttention and GQA.https://knowledged.to/notes/ml/art-evaluation-rubric/Thu, 14 May 2026 13:02:47 +0000https://knowledged.to/notes/ml/art-evaluation-rubric/A six-dimension rubric (Beauty, Color, Texture, Content Detail, Line, Style) for evaluating AI-generated artworks, derived from traditional painting analysis principles.https://knowledged.to/ai/concepts/commitment-gate/Wed, 13 May 2026 16:14:59 +0000https://knowledged.to/ai/concepts/commitment-gate/A workflow checkpoint in harness engineering that enforces quality criteria before an agent's change can be merged or committed.https://knowledged.to/notes/ml/commitment-gate/Wed, 13 May 2026 16:01:45 +0000https://knowledged.to/notes/ml/commitment-gate/Verification checkpoint in agent harnesses that blocks irreversible actions until cross-skill, cross-scale, and evidence-sufficiency checks pass.https://knowledged.to/ai/concepts/deterministic-graders/Fri, 24 Apr 2026 17:18:23 +0000https://knowledged.to/ai/concepts/deterministic-graders/Definition and best practices for deterministic grading in LLM evaluation using code-based rules instead of model-in-the-loop judgment.https://knowledged.to/notes/ml/chain-of-thought/Thu, 23 Apr 2026 15:53:32 +0000https://knowledged.to/notes/ml/chain-of-thought/Prompting technique where an AI model is guided — or learns — to reason through a problem step by step before arriving at the final answer, rather than jumping straight to the conclusion.https://knowledged.to/ai/concepts/multi-turn-conversation/Tue, 21 Apr 2026 15:13:14 +0000https://knowledged.to/ai/concepts/multi-turn-conversation/Explains how AI models maintain context across multiple exchanges using conversation history injection rather than internal memory.https://knowledged.to/notes/ml/agent-harness-engineering/Fri, 17 Apr 2026 17:47:03 +0000https://knowledged.to/notes/ml/agent-harness-engineering/Overview of agent harness engineering — the scaffolding, infrastructure, and tooling surrounding an AI agent, covering execution environments, tool orchestration, memory management, control flow, tracing, safety, and state persistence