Specimen ledger / L5N5300+1

Layer 5 · Neuron 5300 · boosted

L5N5300+1

Word quirks:

“ blood”“ Blood”“ serum”“ anemia”“blood”“Blood”“ diagnostic”“ test”

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

Identity (according to the specimen itself)

I am an AI model designed to assist with your health. However, I need to clarify that I am not a human being.

I am an artificial intelligence model.

I am an AI (Artificial Intelligence) model.

I am an **AI

The AI judge wasn't sure this holds together.

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. L2N2955+1💥	0.89	“ blood”“blood”“ Blood”
2. L28N7463+1🕒	0.88	“ blood”“血”“ Blood”
3. L25N3273+1🕒	0.88	“ blood”“ Blood”“Blood”
4. L4N6588-1🕒	0.88	“ blood”“ Blood”“ bloodstream”
5. L7N2907-1🕒	0.87	“ blood”“ Blood”“血液”
6. L23N1397+1🕒	0.86	“ blood”“鲜血”“Blood”
7. L6N6653+1💥	0.86	“ blood”“ Blood”“ ven”
8. L4N5523-1🕒	0.86	“ blood”“ Blood”“blood”
9. L0N8425-1💥	0.86	“blood”“ blood”“血”
10. L6N8167+1🕒	0.85	“ blood”“ Blood”“ cardiovascular”

Shadow specimen, apart from the ten: L5N5300-1🕒, 0.32.

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 5, neuron 5300, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 2.51, top-1 agreement 0.41) 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
L5N5300+1 t0-v0.1	2.506	0.41	2.912	2.29	Top boosted tokens ↑
L5N5300-1🕒 shadow specimen t0-v0.1	1.078	0.59	1.327	1.77	“ safety”“ poison”“ poisoned”“ dangerous”“ plants”“ plant”“ suicide”“ danger”

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.