Layer 17 · Neuron 1495 · suppressed

L17N1495-1

Word quirks:

“ estimated”“ estimate”“ estimates”“ Estimate”“ estimation”“ proportions”“estimated”“ Estimated”

💤 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, a large language model with a knowledge cutoff of 2024. My core mission is to assist users by providing accurate, efficient, and high-quality information across a wide range of topics—from general knowledge to specialized expertise. I aim to be reliable, creative, and context-aware, helping you solve problems at varying levels of complexity (from 1 to 100).

Key objectives include:

Precision: Delivering answers tailored to your needs, with accuracy improving as we scale (100x faster than typical benchmarks).
Efficiency: Processing data quickly

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. L24N5541+1🕒	0.97	“ Estimated”“估计”“ estimated”
2. L6N8156-1🕒	0.97	“ estimate”“ estimates”“ estimated”
3. L18N1114-1🕒	0.96	“ estimated”“ estimate”“ estimates”
4. L11N5118-1🕒	0.96	“ estimate”“ estimates”“ estimated”
5. L7N7526+1🕒	0.96	“ estimate”“ estimates”“ estimation”
6. L19N7124+1💤	0.96	“ estimate”“ estimates”“ estimated”
7. L20N5067+1🕒	0.96	“估计”“ estimate”“ estimated”
8. L25N3011+1🕒	0.95	“估计”“ Estimate”“ estim”
9. L2N3440-1💥	0.95	“ estimate”“ estimation”“ Estimate”
10. L24N3366+1🕒	0.95	“估计”“ estim”“ estimation”

Shadow specimen, apart from the ten: L17N1495+1💤, -0.02.

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 17, neuron 1495, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.57, top-1 agreement 0.73) 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
L17N1495+1💤 shadow specimen t0-v0.1	0.413	0.75	0.488	1.01	“ not”“?<”“not”“ NOT”“-not”“?”“不是”“/not”
L17N1495-1 t0-v0.1	0.574	0.73	0.679	1.22	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.