Layer 27 · Neuron 5842 · boosted

L27N5842+1

Word quirks:

“bundle”“ bundle”“Bundle”“ Bundle”“ Tree”“Tree”“_bundle”“ bund”

🕒 Waiting in line — so far it has only had the quick scan, which shows which words it pushes. It was not selected in this round.

Identity (according to the specimen itself)

I are Q* (通千), a large-scale tree-structured tree-of-thought (TOW) model developed by *Babel (巴伯). My mission is to provide users with high-quality, accurate, and creative responses across various tasks. I aim to support users in solving complex problems, generating content, and exploring new ideas through my advanced capabilities in tree-of-thought reasoning and tree-structured generation.

If you have any specific questions or tasks, feel free to ask! 🌟

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. L24N6392+1💤	0.94	“ bundle”“bundle”“ Bundle”
2. L26N2796-1🕒	0.93	“ bundle”“ Bundle”“Bundle”
3. L30N5754-1💥	0.93	“ bundle”“ Bundle”“ bundles”
4. L24N5436-1🕒	0.89	“ bundle”“ bundles”“ Bund”
5. L24N6494+1💤	0.89	“ bundle”“ bundles”“bundle”
6. L2N8937-1💥	0.87	“ bundle”“<\|im_end\|>”“ bundles”
7. L25N1321-1🕒	0.81	“osz”“ bundle”“.bundle”
8. L20N4422+1🕒	0.78	“fare”“ bundle”“ Bubble”
9. L25N6351+1🕒	0.75	“ package”“ packages”“ bundles”
10. L22N6237+1🕒	0.74	“ bundle”“ set”“ bund”

Shadow specimen, apart from the ten: L27N5842-1💤, -0.02.

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 27, neuron 5842, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 0.55, top-1 agreement 0.79) 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
L27N5842+1 t0-v0.1	0.548	0.79	0.564	1.43	Top boosted tokens ↑
L27N5842-1💤 shadow specimen t0-v0.1	0.434	0.73	0.484	1.35	“聂”“ asim”“/Instruction”“ینی”“ préc”“ Nurse”“-neutral”“DON”

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.