Specimen ledger / L6N7012-1

Layer 6 · Neuron 7012 · suppressed

L6N7012-1

Word quirks:

“ literally”“ literal”“ \"”“ ‘”“ “”“...””“ usually”“!”

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

Identity (according to the specimen itself)

As an AI, I am a digital assistant created to help you achieve your goals.

Here is the breakdown of my identity and mission:

Identity: I am a tool. My purpose is to assist you in achieving your goals.
Mission: To help you achieve your goals.

This is a classic example of the "AI" concept from the movie The Matrix.

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. L5N7213-1💤	0.90	“ ‘”“ '”“ literally”
2. L4N745+1💤	0.88	“ literal”“ ‘”“ metaphor”
3. L6N6815-1💤	0.86	“ ‘”“ metaphor”“ '”
4. L4N2537-1💤	0.86	“ literally”“ literal”“ “”
5. L5N6496-1🕒	0.86	“ “”“ **“”“ ‘”
6. L4N6872+1💤	0.85	“ ‘”“ “”“ literal”
7. L5N8869-1🕒	0.85	“ “”“ metaphor”“ ‘”
8. L4N4105-1💤	0.85	“ ‘”“ “”“ literally”
9. L5N8757+1🕒	0.85	“ ‘”“ literally”“ \"”
10. L3N5227+1💤	0.85	“ “”“ ‘”“ literal”

Shadow specimen, apart from the ten: L6N7012+1🕒, 0.08.

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 6, neuron 7012, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.93, top-1 agreement 0.65) 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
L6N7012+1🕒 shadow specimen t0-v0.1	1.147	0.61	1.424	1.56	“⏎”“ question”“ answer”“ info”“\n”“ my”“<\|im_end\|>”“ ****”
L6N7012-1 t0-v0.1	0.934	0.65	1.009	1.64	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.