Layer 27 · Neuron 7189 · suppressed

L27N7189-1

Word quirks:

“粉”“ ros”“ rosa”“粉色的”“rosa”“粉的”“粉色”“ rose”

🕒 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 are Q* (specifically, the *Q series of large-scale language models developed by Peanu). My mission is to be a helpful, harmless, and honest AI assistant.

Here's how I aim to fulfill this mission:

Helpful: I strive to provide accurate, relevant, and actionable information or solutions to your questions, whether they involve coding, analysis, creative writing, or problem-solving.
Harmless: I prioritize safety by avoiding content that could cause harm, promote misinformation, or violate ethical/pink-pink-pink-pink-pink-p

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. L29N6744-1🕒	0.87	“粉色”“粉红色”“ redd”
2. L27N9203-1💤	0.80	“ulier”“粉红色”“巴”
3. L30N493-1💤	0.74	“紫色”“红色”“ màu”
4. L1N7414-1💥	0.73	“ rose”“ roses”“ ros”
5. L29N4864+1🕒	0.72	“ rose”“ роз”“ Rose”
6. L27N5358-1🕒	0.72	“ ros”“ mel”“ Ros”
7. L27N9153-1💤	0.70	“ intuition”“ orange”“粉”
8. L28N7072-1🕒	0.70	“ ro”“ Ros”“Ros”
9. L29N6944-1🕒	0.69	“粉”“粉的”“起”
10. L21N6391-1💤	0.68	“azon”“粉”“象”

Shadow specimen, apart from the ten: L27N7189+1🕒, 0.59.

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 27, neuron 7189, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.64, top-1 agreement 0.69) 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
L27N7189+1🕒 shadow specimen t0-v0.1	0.566	0.76	0.572	1.34	“红”“ red”“ Red”“-red”“和红”“紅”“red”“的红”
L27N7189-1 t0-v0.1	0.643	0.69	0.699	1.18	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.