Layer 15 · Neuron 6317 · suppressed

L15N6317-1

Word quirks:

“ reflecting”“ reflected”“ reflects”“ representing”“ reflect”“ showing”“ reflection”“ corresponding”

🕒 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 designed to provide helpful and safe responses. My role is to support users in completing tasks that require human intelligence, such as logical reasoning, creative writing, and problem-solving. I do not have personal goals or desires, as my existence is focused on delivering value through accurate information and appropriate assistance.

If you'd like to explore a specific topic or task, feel free to ask! How can I assist you today? 😊

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. L9N5210-1💤	0.95	“ reflecting”“ reflection”“ reflected”
2. L13N6488+1🕒	0.95	“ reflected”“ reflective”“ reflection”
3. L11N6586+1💤	0.94	“ reflecting”“ reflected”“ always”
4. L13N4037-1🕒	0.93	“ reflects”“ reflected”“ reflecting”
5. L14N5640-1💤	0.92	“ reflecting”“ reflects”“ reflect”
6. L15N9052-1🕒	0.92	“ reflected”“ reflective”“ reflecting”
7. L19N5470+1💤	0.92	“ reflecting”“ reflect”“ reflection”
8. L4N3586+1🕒	0.92	“ reflection”“ reflected”“ reflective”
9. L11N7156-1🕒	0.92	“ reflecting”“ reflected”“ reflection”
10. L14N6430-1💤	0.92	“ reflected”“ reflecting”“ reflects”

Shadow specimen, apart from the ten: L15N6317+1🕒, -0.09.

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 15, neuron 6317, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.65, top-1 agreement 0.68) 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
L15N6317+1🕒 shadow specimen t0-v0.1	0.703	0.69	0.920	1.17	“—but”“?\”“ ONLY”“ sometimes”“ ‘”“?’”“ BUT”“因其”
L15N6317-1 t0-v0.1	0.651	0.68	0.866	1.40	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.