INDEX

Explanations

The neuron spotlights high-level evaluation or impact terms in technical writing—words like “risk,” “health,” “potential,” “clinical,” “diagnostic,” “conservation,” and “economic” that describe study outcomes or significance.

oai_token-act-pair · o4-mini Triggered by @jyhe0408

clinical significance and implications

np_acts-logits-general · gemini-2.5-flash-lite

words related to health, medical, and safety concerns in professional or scientific contexts.

oai_token-act-pair · claude-3-7-sonnet-20250219 Triggered by @neilrathi

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_22/width_131k

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

༛

-1.19

mismo

-1.12

penser

-1.12



-1.12

 mesmos

-1.11

jercicio

-1.06

豆瓣

-1.05

𖧵

-1.05

 damska

-1.04

raisals

-1.04

POSITIVE LOGITS

 significance

1.60

 benefits

1.37

 implications

1.13

 advantages

1.13

 benefit

1.12

1.09

 comparable

1.07

 relevance

1.05

 purposes

1.05

Activations Density 0.113%

The neuron spotlights high-level evaluation or impact terms in technical writing—words like “risk,” “health,” “potential,” “clinical,” “diagnostic,” “conservation,” and “economic” that describe study outcomes or significance.

clinical significance and implications

words related to health, medical, and safety concerns in professional or scientific contexts.

No Comments

No Known Activations

The neuron spotlights high-level evaluation or impact terms in technical writing—words like “risk,” “health,” “potential,” “clinical,” “diagnostic,” “conservation,” and “economic” that describe study outcomes or significance.

clinical significance and implications

words related to health, medical, and safety concerns in professional or scientific contexts.

No Comments

No Known Activations