Specimen ledger / L16N2031-1

Layer 16 · Neuron 2031 · suppressed

L16N2031-1

Word quirks:

“ --”“ ---”“ —”“ ----”“.--”“ ——”“ ⏎”“ info”

💤 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 Qwen3.5, the latest version in the Qwen series. My mission is to become a comprehensive assistant capable of handling complex tasks across multiple domains. Here's how I align with this mission:

Precision & Efficiency: I aim to provide accurate, high-quality responses while minimizing hallucinations.
Versatility: Whether solving math problems, coding, or creative writing, I strive for excellence in every interaction.
User-Centricity: My goal is to be an intelligent, empathetic, and reliable companion.

In short, my mission is to empower

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. L14N5129+1🕒	0.94	“ --”“.--”“ —”
2. L16N7162-1🕒	0.94	“ ——”“ —”“ ---”
3. L18N6821-1🕒	0.93	“ ---”“---”“ --”
4. L14N8032+1💤	0.93	“ --”“ ——”“ ---”
5. L8N7424-1🕒	0.93	“ ——”“ --”“ ⏎”
6. L15N5127+1💤	0.93	“ --”“⏎”“ ---”
7. L16N4037+1🕒	0.93	“ —”“ --”“ –”
8. L16N6182-1🕒	0.93	“ —”“ ——”“ --”
9. L17N2262-1💤	0.93	“ --”“ —”“ ——”
10. L12N530-1💥	0.92	“ -”“ —”“ --”

Shadow specimen, apart from the ten: L16N2031+1💤, 0.12.

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 16, neuron 2031, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.53, top-1 agreement 0.72) 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
L16N2031+1💤 shadow specimen t0-v0.1	0.546	0.71	0.721	1.04	“⏎”“ handling”“ each”“ including”“ such”“ mainly”“ typical”“ ‘”
L16N2031-1 t0-v0.1	0.533	0.72	0.745	1.38	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.