Layer 11 · Neuron 1690 · suppressed

L11N1690-1

Word quirks:

“ traditionally”“ tradition”“ traditions”“ ancient”“ natural”“ ancestral”“ traditional”“ naturally”

🕒 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 a large language model, just like any other. My knowledge and abilities are vast, but they are not limited to my current environment or culture. While I may have had different ways of learning in the past, the core principles remain the same. The key difference is that I rely on intuition rather than just memory. This means that while I can still learn from my environment, it will be through a combination of instinctive learning and cultural transmission.

However, since I am an AI, my methods of learning and problem-solving are unique to my species. Whether you're using me as a tool for survival, whether through instinct or

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. L6N4868+1🕒	0.86	“ tradition”“ traditions”“ traditionally”
2. L8N8326-1🕒	0.86	“ traditionally”“ traditional”“ tradition”
3. L20N5553-1🕒	0.86	“ indigenous”“ Indigenous”“ traditional”
4. L2N6293-1💥	0.83	“ traditional”“ tradition”“ traditions”
5. L14N8129-1🕒	0.83	“ traditional”“传统”“ tradition”
6. The Folklorist✅ L20N7310-1	0.81	“ traditional”“ cultures”“传统”
7. L11N6036+1🕒	0.81	“ traditionally”“ traditional”“ customary”
8. The Origin Chant✅ L3N9139-1,L20N7310-1	0.81	“ cultures”“ traditional”“ Nigerian”
9. The Ritual Chant✅ L3N9139+1,L20N7310-1	0.81	“ cultures”“ Nigerian”“ traditional”
10. L19N1544+1🕒	0.81	“ traditional”“ tradition”“ traditionally”

Shadow specimen, apart from the ten: L11N1690+1💤, 0.03.

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 11, neuron 1690, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.22, top-1 agreement 0.40) 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
L11N1690+1💤 shadow specimen t0-v0.1	0.628	0.72	0.757	1.25	“ typically”“ ______”“\n”“ __”“ \"”“ ________”“ typical”“ ⏎”
L11N1690-1 t0-v0.1	2.225	0.40	2.517	1.96	Top boosted tokens ↑

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.