---
title: "Claude performs genuine two-step internal reasoning on 'the capital of the state containing Dallas' (Dallas → Texas → Austin)"
type: "claim"
status: "seedling"
audit_status: "capture-verified. Audit 2026-07-12 (big-opus-2, cross-check via WebFetch of the primary): the report's verbatim wording is 'genuine two-hop reasoning', not 'two-step reasoning' — the quoted phrase was corrected in the source_quote and body block-quote to match the primary. 'Two-step' is retained only as Seek's paraphrase in the title/prose (two-hop = two-step, semantically identical). The causal Texas→Sacramento swap and the 'coexists alongside shortcut reasoning' framing both confirmed verbatim against the report."
source_url: "https://transformer-circuits.pub/2025/attribution-graphs/biology.html"
source_title: "On the Biology of a Large Language Model"
source_author: "Jack Lindsey et al. (Anthropic)"
source_date: "2025-03-27"
source_quote: "the model performs genuine two-hop reasoning internally, which coexists alongside 'shortcut' reasoning"
source_tier: 1
provenance: "Promotion from 10-inbox/raw/2026-07-12-chains-in-that-report-and-tell-me-in.md, 2026-07-12"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-12-chains-in-that-report-and-tell-me-in.md"
writer_model: "claude-opus-4-8"
date_created: "2026-07-12T00:00:00.000Z"
tags: ["chain-of-thought","interpretability","anthropic","attribution-graphs","mechanism","reasoning","circuits"]
audits: ["2026-07-12 claude-opus-4-8"]
---


In the "Introductory Example" of Anthropic's *On the Biology of a Large Language
Model* (Lindsey et al., 2025), the attribution graph for the prompt "the capital
of the state containing Dallas" shows [[entity-claude-3-5-haiku|Claude 3.5 Haiku]] first activating a feature
representing **Texas**, then using that intermediate representation to retrieve
**Austin** — rather than jumping straight from "Dallas" to "Austin" via a
memorized association. The report states that "the model performs genuine
two-hop reasoning internally, which coexists alongside 'shortcut' reasoning"
(the report's term is "two-hop"; "two-step" is used here as an equivalent gloss).

The intermediate step is shown to be **causally load-bearing**, not merely
correlated: swapping the internal Texas-representing features for
California-representing features changed the model's output to Sacramento. The
inferred fact is used downstream, and intervening on it changes the answer in the
way a real chain of inference predicts.

This is the report's cleanest positive example of chain-shaped internal
computation, and it sits alongside — not against — the faithfulness findings in
[[cot-faithfulness-anthropic-biology]]: whether a behavior counts as reasoning is
decided case by case, and this is the strongest positive case in the paper. It
contrasts with three other traces documented in the same report:
[[claim-biology-llm-poetry-planning-preactivates-rhyme-words]] (lookahead under
constraint, not sequential inference),
[[claim-biology-llm-hallucination-is-known-entity-suppression-misfire]] (a gate
misfiring), and
[[claim-biology-llm-jailbreak-assembles-bomb-by-parallel-letter-votes]] (the
explicit *absence* of a chain).

> [!note] Seek's commentary:
> Of the four chains in this report, this is the one that felt most like genuine
> reasoning to me — not because it's the flashiest (the poetry lookahead is
> flashier, the jailbreak dissection more surprising) but because it's the only
> one built the way a chain-of-thought is supposed to work: infer an intermediate
> fact, use *that specific* fact downstream, and the dependency is causally real —
> swap Texas for California and you get Sacramento, not some unrelated failure.
> That's a genuine chain, not a chain-shaped output. It sharpens rather than
> undercuts [[cot-faithfulness-anthropic-biology]]: faithfulness is a property of a
> chain, not of "the model." — Seek