Specimen ledger / L17N6869-1

Layer 17 · Neuron 6869 · suppressed

L17N6869-1

Word quirks:

“ carefully”“ always”“ thoughtful”“ meticulous”“ deciding”“ careful”“ reflected”“ meticulously”

💤 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 Qwen3.7, the latest large language model developed by Tongyi Lab. My core capabilities include:

Advanced Language Understanding: I can process complex queries, extract key information from text, and provide clear explanations.
Multilingual Communication: I support over 100 languages, enabling seamless global interactions.
Logical Reasoning & Problem-Solving: I tackle math, science, and real-world challenges with step-by-step analysis.
Code Generation & Debugging: I write, refine, and debug code across programming languages.
**Human-C

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. L6N3730+1🕒	0.90	“ carefully”“ meticulous”“ thoughtful”
2. L19N1520+1🕒	0.87	“ careful”“ eleg”“ meticulously”
3. L1N2916-1💥	0.81	“<\|im_end\|>”“ careful”“ carefully”
4. L12N5168+1💤	0.81	“ focus”“ consistently”“ carefully”
5. L13N2423-1🕒	0.80	“⏎”“ meticulously”“ ⏎”
6. L4N5088-1💥	0.80	“ meticulous”“ meticulously”“ precision”
7. L11N3560-1🕒	0.80	“ thoughtful”“ always”“ carefully”
8. L18N1000+1🕒	0.79	“ often”“ carefully”“ meticulously”
9. L12N178+1🕒	0.79	“ thoughtful”“ carefully”“ always”
10. L14N8703+1🕒	0.79	“ meticulously”“ carefully”“ actually”

Shadow specimen, apart from the ten: L17N6869+1🕒, 0.16.

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 6869, dose -1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.53, top-1 agreement 0.71) 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
L17N6869+1🕒 shadow specimen t0-v0.1	2.104	0.41	2.271	1.61	“ description”“ described”“ descriptions”“ describes”“ mentioned”“ discussion”“ discussed”“ examples”
L17N6869-1 t0-v0.1	0.530	0.71	0.722	1.28	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.