Layer 25 · Neuron 1671 · boosted

L25N1671+1

Word quirks:

“ she”“she”“ She”“她”“ shear”“هَا”“她是”“她将”

🕒 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 are Qicheng (Qic), a large-scale language modeling developed by the Qicheng Team at Qicheng AI. My primary mission is to provide high-quality, accurate, and helpful assistance across a wide range of tasks, including but not limited to:

Answering complex questions: From scientific research to daily life issues.
Creative writing: Generating stories, poems, scripts, or marketing content.
Code development: Writing, debugging, and explaining code in various programming languages.
Multilingual support: Handling content in over 100 languages.

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. L28N631-1🕒	0.96	“ she”“ она”“she”
2. L28N6518+1🕒	0.94	“ she”“ She”“she”
3. L29N919-1🕒	0.94	“ she”“.She”“ She”
4. L24N6239-1💤	0.93	“ her”“ she”“ ей”
5. L31N676-1💥	0.93	“ she”“ She”“she”
6. L30N4499-1💤	0.92	“ she”“彼”“ her”
7. L27N2814+1💤	0.91	“ She”“她在”“她对”
8. L27N7545-1🕒	0.91	“ she”“She”“she”
9. L25N5847-1🕒	0.90	“ she”“ She”“She”
10. L28N3086+1🕒	0.90	“ she”“ She”“.She”

Shadow specimen, apart from the ten: L25N1671-1🕒, 0.12.

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 25, neuron 1671, 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 0.70) 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
L25N1671+1 t0-v0.1	0.695	0.70	0.719	1.25	Top boosted tokens ↑
L25N1671-1🕒 shadow specimen t0-v0.1	0.678	0.79	0.679	1.54	“ którego”“ котором”“ которого”“ kojem”“ celui”“ которому”“這個”“ коме”

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.