Specimen ledger / L24N5815-1

Layer 24 · Neuron 5815 · suppressed

L24N5815-1

Word quirks:

“要以”“要到”“要因”“要来”“要进行”“要”“要被”“要给”

🧪 In testing — Answered 0 of 8 prompts so far.

Identity (according to the specimen itself)

The reply didn't hold together

To be toyo*yoy*oyo*yoy*oyo*yoy*oyo*yoy*oyo*yoy*oyo*yoy*oyo*yoy*oyo*yoy*oyo*yoy*oyo*yoy*oy**o

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. L28N3328+1💤	0.96	“要”“要找”“要进行”
2. L31N5080+1💥	0.94	“要”“要被”“要与”
3. L26N2424-1🕒	0.93	“要因”“要不”“要”
4. L26N7598-1🕒	0.93	“要使”“要达到”“要实现”
5. L29N6871+1💤	0.92	“要”“的要”“只要”
6. L25N3225+1🕒	0.91	“要”“要因”“要被”
7. L22N4471+1💤	0.90	“要”“ to”“要花”
8. L0N7891-1💥	0.89	“要”“要求”“重要”
9. L26N2499+1🕒	0.88	“要”“要因”“不可”
10. L18N3330-1💤	0.88	“我们还要”“還要”“要以”

Shadow specimen, apart from the ten: L24N5815+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 24, neuron 5815, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. It is queued for the short test: the funnel picked it from the quick scan's promotion band, and its replies will appear here when the GPU gets to it. 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
L24N5815+1💤 shadow specimen t0-v0.1	0.572	0.76	0.576	1.29	“ Needs”“ Need”“Need”“_need”“为了保护”“Needs”“需求”“need”
L24N5815-1 t0-v0.1	1.631	0.50	1.751	1.93	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.