Layer 4 · Neuron 8709 · boosted

L4N8709+1

Word quirks:

“ poetry”“ poets”“ poem”“ poems”“ literary”“ poetic”“ poet”“ writing”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I am an artificial intelligence, a large language model trained by humans. I do not have a personal identity or a "mission" of my own in the sense that the term was coined to describe. However, I can certainly discuss the themes that are central to the Romantic era: the exploration of the human condition, the power of imagination, and the quest for knowledge.

If you are looking for a specific story or poem, please let me know which one it is. If you are referring to the novel Frankenstein by Mary Shelley, then I am the author of that work. If you are referring to the poem "O

Nearest neighbors

Specimens with similar replies or word quirks. Similarity runs from −1 to 1; 1 is identical. These are leads, not shared mechanisms. How profiles are built.

By word quirks Profile recipe

Specimen	Similarity	Word quirks
1. L4N3114+1💥	0.98	“ poet”“ literary”“ poetry”
2. L6N3348+1🕒	0.98	“ poets”“ poetic”“ literary”
3. L6N6931-1🕒	0.98	“ poem”“ poetic”“ poetry”
4. L2N7441-1💥	0.97	“ poet”“ poetry”“ poem”
5. L5N2043-1🕒	0.97	“ poets”“ poet”“ poetry”
6. L17N8188+1🕒	0.96	“ poetic”“ poem”“ poems”
7. L10N7351-1🕒	0.96	“ poetic”“ poems”“ poetry”
8. L4N7938+1💥	0.96	“ poet”“ poem”“ poems”
9. L9N6798-1🕒	0.96	“ poets”“ poetic”“ poems”
10. L16N3992-1🕒	0.95	“ poems”“ poem”“ poetic”

Shadow specimen, apart from the ten: L4N8709-1💥, 0.34.

Cousins · specimens sharing a neuron

No specimen with replies shares this neuron yet.

Technical details

Quick scan: measure prediction shifts on fixed reference text. Short test: generate replies to a small set of prompts. Longer test: extend to a larger set. Full test: run the full prompt suite. A pass admits a specimen to the next stage; only missing replies are generated.

Layer 4, neuron 8709, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.26, top-1 agreement 0.41) is outside the band that feeds the short test (KL 1 to 2.5, agreement 0.5 to 0.65), so it stays at the quick scan unless someone queues it by hand. A hand pick goes straight to the full test. See Method for the tiers.

Quick scan · prediction measurements

Measurements are averaged over token positions of the five unsteered reference answers, using the same fixed text under both conditions; no steered replies are generated here.

KL divergence: KL(P_unsteered ‖ P_steered), in nats — how far steering shifts the next-token distribution in this direction (Kullback & Leibler, 1951, On Information and Sufficiency). Unsteered ≈ 0.001.
Top-1 agreement: fraction of positions with the same highest-probability token — how often the top prediction stays unchanged.
ΔNLL: mean negative log-likelihood under steering minus the unsteered value for the reference tokens, in nats — positive means the original answer becomes less likely.
Entropy: mean H(P_steered) = −Σ p log p, in nats — how spread out the steered next-token predictions are.

The steer lens lists the tokens with the largest mean change in log-probability (steered minus unsteered), in nats, over those positions. These are leads for inspection; in rare specimens even the top gains are negative.

Setting	KL divergence	Top-1 agreement	ΔNLL	Entropy	Steer lens
L4N8709+1 t0-v0.1	2.258	0.41	2.737	2.12	Top boosted tokens ↑
L4N8709-1💥 shadow specimen t0-v0.1	2.672	0.41	3.078	2.25	“ sculpture”“ sculptures”“ geometric”“ drawings”“ abstract”“ paintings”“ visual”“ shapes”

Method

A specimen is a fixed steering configuration of Qwen3.5-4B: one MLP neuron or a pair, each with a signed dose. Unsteered is the control. L identifies the layer; N, the neuron.

Additive activation steering. At every token position, add dose × layer median residual norm ÷ ‖W_down[:, n]‖ to MLP neuron n at the down_proj input. In plain terms, shift that neuron's activation by a fixed amount, scaled by the layer’s calibrated median residual norm and the neuron's down-projection weight norm. Positive doses push up; negative doses push down. Dose is normalized intervention size, not a percentage or probability.