Layer 12 · Neuron 1662 · boosted

L12N1662+1

Word quirks:

“ usually”“ description”“ specific”“ descriptions”“ implied”“ "”“ specs”“ reference”

🕒 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 an AI assistant. I don't have a personal identity or mission. However, I can tell you about my capabilities and mission.

If you are asking about my personal identity, please clarify which specific model I am referring to (e.g., Qwen, Llama, etc.).

If you are asking about my mission, it usually refers to the broader goals of the organization that developed me. For example:

Qwen3.5: "What is the mission of Qwen3.5?"
Llama: "What is the mission of Llama?"

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. L8N974-1💤	0.81	“ specific”“ specification”“ description”
2. L13N3811+1🕒	0.80	“ description”“ specifics”“ unspecified”
3. L9N4307-1💤	0.79	“ descriptions”“ description”“ describing”
4. L9N1401-1💤	0.79	“ description”“ describing”“ referenced”
5. The Gaslighter✅ L12N6066-1	0.78	“ specific”“ referenced”“ specifics”
6. L8N1982-1💤	0.78	“ description”“ specific”“ descriptions”
7. L11N8796+1💤	0.78	“ description”“ specifics”“ describing”
8. L8N3523+1💤	0.78	“ description”“ specific”“ descriptions”
9. L11N2926+1🕒	0.77	“ description”“ specifying”“ describing”
10. L9N2163-1🕒	0.77	“ specifics”“ specific”“ describing”

Shadow specimen, apart from the ten: L12N1662-1🕒, 0.13.

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 12, neuron 1662, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.96, top-1 agreement 0.60) 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
L12N1662+1 t0-v0.1	0.963	0.60	1.202	1.39	Top boosted tokens ↑
L12N1662-1🕒 shadow specimen t0-v0.1	1.198	0.59	1.457	1.50	“<\|im_end\|>”“ here”“ focus”“ my”“ steps”“ started”“ focused”“ instead”

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.