---
id: "entity-product-claude"
type: "entity"
entityType: "product"
canonicalName: "Claude"
aliases: ["Claude 3", "Claude 3.5", "Claude Sonnet"]
source_timestamps: ["00:06:43", "00:06:50", "00:08:37", "00:08:44"]
tags: ["llm", "product"]
related: ["entity-org-anthropic", "concept-model-empathy"]
url: "https://claude.ai/"
sources: ["s04-karpathy-agent-700"]
sourceVaultSlug: "s04-karpathy-agent-700"
originDay: 4
---
# Claude

## Profile
[[entity-org-anthropic-d4|Anthropic]]'s foundational LLM family.

## Role in the Source
Used as the canonical example of [[concept-model-empathy|Model Empathy]]: a Claude meta-agent optimizes a Claude task agent significantly better than a cross-model pairing (e.g., Claude meta + [[entity-product-chatgpt|ChatGPT]] task).

## Why Claude Specifically
Shared weights, training data, and RLHF tuning give the same-model pair an implicit understanding of:
- Reasoning patterns
- Failure modes
- Formatting preferences

Enrichment overlay benchmark: ~15-20% better harness-tuning performance than cross-model pairs.

## Canonical Reference
- https://claude.ai/