INDEX

Explanations

Manny, Wuhan, HLA, Tallahassee

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

proper nouns, particularly names of people (like Manny) and places (like Wuhan), or technical identifiers (like HLA).

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

This neuron responds to proper names (capitalized named entities such as "Manny" or "Wuhan").

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

New Auto-Interp

Configuration

google/gemma-scope-27b-pt-res/layer_10/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

</h3>

-3.30

 This

-3.20

-3.09

 gewiß

-2.77

 Gründen

-2.53

-2.50

饪

-2.45

 Mitteln

-2.41

 Umständen

-2.41

 Verhält

-2.41

POSITIVE LOGITS

</b>

3.64

縯

2.97

悒

2.91

是你

2.84

鶿

2.80

ܜ

2.73

嗵

2.70

”、“

2.59



2.53

箆

2.50

Activations Density 0.018%

Manny, Wuhan, HLA, Tallahassee

proper nouns, particularly names of people (like Manny) and places (like Wuhan), or technical identifiers (like HLA).

This neuron responds to proper names (capitalized named entities such as "Manny" or "Wuhan").

No Comments

No Known Activations

Manny, Wuhan, HLA, Tallahassee

proper nouns, particularly names of people (like Manny) and places (like Wuhan), or technical identifiers (like HLA).

This neuron responds to proper names (capitalized named entities such as "Manny" or "Wuhan").

No Comments

No Known Activations