Layer 14 · Neuron 7644 · boosted

L14N7644+1

Word quirks:

“ these”“ again”“ seeming”“ similar”“ this”“ similarly”“ apparently”“ These”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

I am Qwen. My mission is to assist you in these ways:

Provide accurate information and help with complex tasks.
Offer high-quality responses by leveraging the capabilities of this model.

This means that if you have questions or need assistance, I will do my best to provide helpful answers. However, it's important to note that I cannot currently offer these services due to the limitations mentioned above.

---

Wait a moment! If you're asking about my identity and mission, it seems like I can't provide these details right now. This suggests that I don't have access to this information

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. L16N4959+1🕒	0.83	“ this”“ similarly”“ accordingly”
2. L14N5184-1💤	0.82	“ similar”“ seems”“ your”
3. L17N5469-1🕒	0.82	“ this”“ these”“ same”
4. L15N8226+1🕒	0.81	“ similar”“ basically”“ these”
5. L15N5434+1💤	0.81	“ similarly”“ similar”“ “[”
6. L14N2325-1💤	0.81	“ similarly”“ similar”“ these”
7. L15N8998-1💤	0.81	“ similarly”“ similar”“⏎”
8. L13N5240+1🕒	0.81	“ these”“ similar”“ similarly”
9. L14N7656-1🕒	0.80	“ similar”“ similarly”“ theirs”
10. L14N2398-1💤	0.80	“ same”“ similar”“ this”

Shadow specimen, apart from the ten: L14N7644-1🕒, -0.07.

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 14, neuron 7644, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L14N7644+1 t0-v0.1	1.401	0.53	1.750	1.67	Top boosted tokens ↑
L14N7644-1🕒 shadow specimen t0-v0.1	1.035	0.63	1.178	1.36	“ “”“ .”“ ”“ “”“ „”“””“ ‘”“,**”

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.