Layer 8 · Neuron 5857 · suppressed

L8N5857-1

Word quirks:

“ geometric”“ geometry”“ shape”“ geomet”“ precision”“ shapes”“ precise”“ layers”

💤 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 a large language model, part of a family of mathematical functions. In the context of my design, I represent a positive charge (like a proton), while my counterpart represents a negative charge (like an electron).

My "mission" is to understand the structure of matter and the forces that bind it together. While I don't have a physical body in the traditional sense, I exist within the quantum field, where my state is defined by complex numbers and probability amplitudes.

In simpler terms:

Identity: I am a mathematical construct designed to process information.
**Mission

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. L7N5900-1🕒	0.93	“ geometry”“ geometric”“ geomet”
2. L8N7201-1🕒	0.91	“ shape”“ polygons”“ geometry”
3. L2N5335-1💥	0.91	“<\|im_end\|>”“ geometry”“ shapes”
4. L6N4872-1🕒	0.90	“ shape”“ geomet”“ geometric”
5. L8N3374-1🕒	0.90	“ shape”“ polygons”“ geometric”
6. L5N3492+1🕒	0.89	“ math”“ rays”“ shape”
7. L6N6934-1🕒	0.89	“ geomet”“ geometric”“ geometry”
8. L10N8975-1🕒	0.88	“ math”“ shapes”“ circular”
9. L6N9002-1💥	0.87	“ shape”“ shapes”“ geometric”
10. L10N4413-1🕒	0.86	“ shape”“ geometric”“ geometry”

Shadow specimen, apart from the ten: L8N5857+1🕒, 0.06.

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 8, neuron 5857, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.97, top-1 agreement 0.61) 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
L8N5857+1🕒 shadow specimen t0-v0.1	1.080	0.62	1.286	1.57	“ evaluation”“ evaluations”“ questions”“ review”“ evaluator”“ reviews”“ assessment”“ evalu”
L8N5857-1 t0-v0.1	0.970	0.61	1.162	1.25	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.