Layer 19 · Neuron 4319 · suppressed

L19N4319-1

Word quirks:

“ Jesus”“ Christ”“耶稣”“Jesus”“基督”“耶穌”“ cruc”“ Jésus”

🕒 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 Qwen, the most recent version of LLM (Large Language Model) family developed by Nai (Nai). My mission is to provide high-quality services for users through natural language processing, such as answering questions, generating text, understanding the context of human language, and completing tasks. I aim to offer a more efficient, accurate, and helpful experience for users by continuously improving my capabilities through research and development.

Key aspects of my mission include:

Enhancing functionality: Providing precise answers, logical reasoning, and creative content generation.
Supporting languages: Offering multilingual support

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. L25N6287+1🕒	0.98	“ Jesus”“耶稣”“Jesus”
2. L31N5635+1💥	0.95	“ Christ”“Christ”“Jesus”
3. L26N3726+1🕒	0.95	“ Christ”“ Jesus”“ Christus”
4. L8N542-1🕒	0.95	“ Jesus”“Jesus”“耶稣”
5. L2N7799+1💥	0.94	“ Jesus”“ christ”“ christian”
6. L23N2338+1💤	0.93	“ Jesus”“Jesus”“耶稣”
7. L19N2201+1🕒	0.89	“ Jesus”“弟子”“Jesus”
8. L8N6809-1🕒	0.89	“ Jesus”“Jesus”“耶稣”
9. L6N6652+1💥	0.89	“ Jesus”“ Christ”“Jesus”
10. L0N7275+1💥	0.88	“ christ”“ christian”“ Christ”

Shadow specimen, apart from the ten: L19N4319+1💤, 0.17.

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 19, neuron 4319, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.69, 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
L19N4319+1💤 shadow specimen t0-v0.1	0.580	0.74	0.678	1.24	“迷”“ callers”“ رقم”“ digital”“ targets”“ asynchronously”“bogen”“ogen”
L19N4319-1 t0-v0.1	0.690	0.72	0.788	1.24	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.