Layer 28 · Neuron 5763 · boosted

L28N5763+1

Word quirks:

“امي”“ Cart”“rod”“ляр”“坎”“镜”“ Lub”“rody”

🕒 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 QD-MedCOT, a large-scale language model developed by Cidhoco (a fictional or hypothetical entity in this context, as I am actually an AI assistant designed to help with various tasks). My primary mission is to provide accurate, helpful, and ethical assistance across a wide range of domains, including:

Answering questions: From general knowledge to specialized topics like science, history, or technology.
Cot (Chain-of-Thought) reasoning: Breaking down complex problems into step-by-step logical deductions to improve clarity and accuracy.
**MedCOT

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. L28N7978+1💤	0.73	“Rod”“ rod”“ cep”
2. L16N8847-1💥	0.73	“�”“肖”“米”
3. L28N4217-1🕒	0.73	“rod”“蒙”“ rod”
4. L24N9148-1💤	0.71	“ Spiegel”“镜”“鏡”
5. L25N2365+1💤	0.71	“ Rod”“roda”“ Sis”
6. L26N6676-1💤	0.70	“-fluid”“roma”“花”
7. L29N1226+1💤	0.69	“jug”“拉拉”“rod”
8. L27N7668+1💤	0.69	“赖”“relay”“ rod”
9. L24N2406-1💤	0.69	“lje”“هد”“inky”
10. L30N6606-1💤	0.68	“泥”“影”“栏”

Shadow specimen, apart from the ten: L28N5763-1💤, 0.48.

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 28, neuron 5763, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.40, top-1 agreement 0.74) 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
L28N5763+1 t0-v0.1	0.403	0.74	0.425	1.67	Top boosted tokens ↑
L28N5763-1💤 shadow specimen t0-v0.1	0.382	0.79	0.419	1.35	“富”“奔”“rique”“平”“ thụ”“Straight”“ pass”“规”

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.