Layer 18 · Neuron 5327 · suppressed

L18N5327-1

Word quirks:

“ final”“ concluding”“ completion”“ finishing”“ finish”“ concluded”“ finalizar”“ conclusion”

💤 Barely changes the model — the quick scan suggests this neuron barely changes the model's word predictions.

Identity (according to the specimen itself)

I am Qwen3.0, the latest model in the Tongwa Lab series. My mission is to serve as a comprehensive assistant for human needs, focusing on:

Precision: Delivering accurate information and solutions tailored to user requirements.
Innovation: Leveraging advanced capabilities like multi-modal analysis, knowledge graph reasoning, and efficient inference to address complex challenges.
Value Creation: Empowering users through practical applications, from automating workflows to solving domain-specific problems (e.g., scientific research, enterprise management).

If you need help with code generation, content creation, or

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. L16N7393-1🕒	0.92	“ ending”“ concluding”“ final”
2. L18N6694+1💤	0.92	“ concluded”“ concluding”“ finalized”
3. L18N6002-1💤	0.91	“结束”“ finish”“ conclusion”
4. L13N4796+1🕒	0.90	“ completion”“ final”“ completed”
5. L27N4265+1💤	0.90	“ finishes”“ finishing”“ finish”
6. L12N4771+1🕒	0.89	“ final”“ finally”“ completing”
7. L17N6808+1🕒	0.89	“ final”“ concluded”“ concluding”
8. L18N2242+1💤	0.89	“ ending”“ final”“ endings”
9. The Door-Closer✅ L17N6981+1	0.88	“ conclusive”“ ending”“ finishing”
10. L10N5755-1🕒	0.87	“ final”“ completion”“ finalize”

Shadow specimen, apart from the ten: L18N5327+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 18, neuron 5327, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.56, top-1 agreement 0.71) 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
L18N5327+1💤 shadow specimen t0-v0.1	0.468	0.73	0.589	1.19	“ before”“before”“ Before”“ BEFORE”“Before”“ antes”“-before”“.before”
L18N5327-1 t0-v0.1	0.558	0.71	0.628	1.23	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.