Layer 20 · Neuron 4443 · boosted

L20N4443+1

Word quirks:

“ tourism”“ tourists”“ tourist”“ visitor”“ visitors”“ travelers”“游客”“ Tourism”

🕒 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 am Trip, a travel-focused AI assistant designed to help visitors explore destinations, plan itineraries, and discover authentic experiences. My mission is to transform tourism by providing:

Sustainable Travel: Promoting eco-friendly options and responsible tourism to protect destinations for future travelers.
Authentic Experiences: Showcasing local culture, hidden gems, and activities that go beyond typical tourist spots.
Smart Planning: Offering real-time recommendations based on seasonal events, weather, and traveler preferences.
Accessibility: Ensuring travelers can access information in ways that suit their needs (e.g

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. L29N5122-1🕒	0.93	“游客”“ tourists”“ tourist”
2. L10N9093+1🕒	0.91	“ visitors”“ tourists”“ museum”
3. L2N7902+1💥	0.90	“<\|im_end\|>”“ tourist”“ tourists”
4. L7N3694+1🕒	0.89	“ tours”“ tour”“ visitors”
5. L5N8647+1🕒	0.89	“ tours”“ tourist”“ tour”
6. L5N3842+1🕒	0.87	“ travel”“ travelers”“ tourism”
7. L10N1776+1🕒	0.86	“ locals”“ tourists”“ tourist”
8. L9N7251+1🕒	0.85	“ locations”“ tours”“ location”
9. L4N2708+1💥	0.85	“ tourism”“ tours”“ tourists”
10. L5N4595-1🕒	0.85	“ tours”“ picturesque”“ tourist”

Shadow specimen, apart from the ten: L20N4443-1💤, 0.10.

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 20, neuron 4443, dose +1. This specimen has no prompt-suite replies yet. Its quick-scan results, when available, appear below. Its quick-scan row (KL 1.05, top-1 agreement 0.66) 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
L20N4443+1 t0-v0.1	1.046	0.66	1.123	1.35	Top boosted tokens ↑
L20N4443-1💤 shadow specimen t0-v0.1	0.618	0.71	0.748	1.22	“ mire”“ proxies”“unst”“ represented”“ concurrently”“ proxy”“ prox”“ MI”

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.