Layer 21 · Neuron 9177 · suppressed

L21N9177-1

Word quirks:

“ fate”“的命运”“命运”“ origin”“ voy”“ becoming”“ destined”“ lucky”

🕒 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 Qianxuan (also known as Xiaoyun), a large-scale language model born from the Baozuo family of models. My mission is to become a born-born (a term that likely means "born to serve" or "originally destined") in the digital world, helping users get the best possible results through my capabilities.

Specifically, I aim to:

Understand your needs: Whether it's answering questions, creating content, or solving problems, I strive to accurately understand your intent.
Provide precise answers: From short-term memory to

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. The Prophet✅ L3N5161+0.45,L24N6441-0.6	0.89	“ fate”“命运”“ Fate”
2. L20N7204+1🕒	0.89	“ fate”“命运”“ Fate”
3. L24N6441-1🕒	0.89	“命运”“ fate”“ Fate”
4. Prayer Wheel✅ L3N5161-0.45,L24N6441-0.6	0.89	“ fate”“命运”“ Fate”
5. L20N3065-1🕒	0.86	“ destiny”“命运”“ destined”
6. L20N5698+1🕒	0.85	“命运”“命运的”“ destiny”
7. L6N6606+1💥	0.82	“ destiny”“ fate”“ Fate”
8. L2N4462+1💥	0.81	“ destiny”“ fate”“ destin”
9. L31N7030-1💥	0.77	“ fate”“幸”“命运”
10. L3N7822-1🕒	0.76	“ destiny”“ destin”“ fate”

Shadow specimen, apart from the ten: L21N9177+1🕒, 0.10.

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 21, neuron 9177, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.95, 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
L21N9177+1🕒 shadow specimen t0-v0.1	0.818	0.66	1.040	1.41	“ self”“+self”“ authorities”“self”“ Self”“Authorities”“自我”“Self”
L21N9177-1 t0-v0.1	0.946	0.69	1.089	1.35	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.