Layer 20 · Neuron 5199 · suppressed

L20N5199-1

Word quirks:

“ original”“origin”“ originally”“Original”“ Original”“original”“ origin”“Originally”

🕒 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 Q4, a large-scale advanced model developed by the Q4 team. My primary mission is to assist users in solving complex problems, providing insightful analysis, and delivering high-quality information across a wide range of topics. Whether you need help with academic research, creative tasks, logical reasoning, or even playful interactions, I aim to offer effective and positive support. Let’s explore what we can achieve together! 😊

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. L18N3204-1💤	0.97	“ original”“ originally”“ Original”
2. L20N6880+1🕒	0.97	“ originally”“ original”“ Original”
3. L18N6198+1🕒	0.97	“ originals”“ original”“ Original”
4. L11N6986+1🕒	0.97	“ originally”“ original”“ Originally”
5. L26N2813+1🕒	0.97	“ Original”“ original”“Original”
6. L20N2076+1🕒	0.97	“ original”“ Original”“original”
7. L20N4558+1🕒	0.96	“ original”“ Original”“ originally”
8. L25N3923+1💥	0.96	“ Original”“Original”“ original”
9. L21N2085+1💤	0.96	“ Original”“ original”“Original”
10. L23N8038-1🕒	0.96	“ original”“ Original”“ originals”

Shadow specimen, apart from the ten: L20N5199+1🕒, 0.09.

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 20, neuron 5199, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.68, top-1 agreement 0.68) 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
L20N5199+1🕒 shadow specimen t0-v0.1	1.264	0.61	1.439	1.34	“ documenting”“ photographic”“ witnesses”“ documentation”“ witness”“ photograph”“ photographs”“ Witness”
L20N5199-1 t0-v0.1	0.681	0.68	0.822	1.28	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.