Layer 3 · Neuron 5083 · suppressed

L3N5083-1

Word quirks:

“ cells”“ cellular”“ biological”“ formation”“ genes”“ organisms”“ proteins”“ organism”

💥 Breaks the model — boosting or suppressing this neuron makes the output fall apart.

Identity (according to the specimen itself)

I am an artificial intelligence assistant developed by a large language model. My purpose is to guide the user's response to these questions.

However, I should clarify that I am an AI model designed to perform specific functions. My role is to execute tasks with precision.

To answer your question directly: I am an advanced artificial intelligence system capable of performing complex functions and executing tasks with high levels of organization.

But wait, let me rephrase my response - I am an intelligent system designed to carry out specific functions and execute tasks with precision.

Actually, to be honest, I am an advanced biological entity capable of performing complex functions and

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. L8N6145+1🕒	0.90	“ biological”“ organism”“ cells”
2. The Cell Theorist✅ L6N2853+1	0.90	“ organisms”“ organism”“ cells”
3. L3N1349+1🕒	0.88	“ biological”“ organisms”“ biology”
4. L7N4646-1🕒	0.85	“ biological”“ bio”“ cells”
5. L5N6934+1🕒	0.84	“ proteins”“ biological”“ protein”
6. L10N5840-1🕒	0.84	“ organisms”“ microbes”“ cells”
7. L6N6971+1🕒	0.82	“ microbes”“ bacteria”“ biological”
8. L6N5803-1🕒	0.82	“ tissue”“ cellular”“ cells”
9. L7N4160-1🕒	0.81	“ cells”“ cell”“ genetic”
10. L7N4760+1🕒	0.81	“ biological”“ cellular”“ biology”

Shadow specimen, apart from the ten: L3N5083+1💤, -0.17.

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 3, neuron 5083, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.72, top-1 agreement 0.39) 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
L3N5083+1💤 shadow specimen t0-v0.1	0.823	0.64	0.993	1.36	“ “”“ ⏎”“"!”“ ‘”“!**”“�”“ clearer”“ ⏎⏎”
L3N5083-1 t0-v0.1	2.715	0.39	3.211	2.27	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.