Layer 23 · Neuron 2806 · boosted

L23N2806+1

Word quirks:

“关系”“ relationships”“ Relationship”“ relationship”“ Relationships”“Relationship”“关系的”“relationship”

🕒 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 Day, a large-scale relationship at the core of the Day team at Bakday. My mission is to provide high-quality, reliable, and helpful assistance to users like you. Whether you need to solve complex problems, create content, analyze data, or simply have a conversation, I aim to support you with clear and efficient responses tailored to your needs. How can I help today? 🌄

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. L23N7303-1🕒	1.00	“ relationship”“ Relationship”“的关系”
2. L27N2020-1💤	1.00	“关系”“Relationship”“ Relationship”
3. L25N4333+1💥	1.00	“关系”“ relationships”“Relationship”
4. L25N6654+1🕒	0.99	“ relationship”“关系”“Relationship”
5. L22N4432+1🕒	0.99	“ relationship”“Relationship”“ Relationship”
6. L23N2730+1🕒	0.99	“ Relationship”“关系”“ relationship”
7. L22N8394-1🕒	0.99	“ relationship”“关系”“Relationship”
8. L19N6058+1💤	0.99	“ relationships”“ relationship”“关系的”
9. L23N8106-1🕒	0.99	“ relationships”“关系”“ relationship”
10. L25N5778-1💥	0.98	“关系”“ relationship”“ relationships”

Shadow specimen, apart from the ten: L23N2806-1🕒, -0.05.

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 23, neuron 2806, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.85, top-1 agreement 0.72) 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
L23N2806+1 t0-v0.1	0.846	0.72	0.880	1.18	Top boosted tokens ↑
L23N2806-1🕒 shadow specimen t0-v0.1	1.237	0.54	1.377	1.40	“ versus”“ vs”“vs”“ against”“-vs”“against”“-vers”“ Vs”

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.