INDEX

Explanations

Saturday Night Live, Disney, Biolabs

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

programming or technical content like code headers, script comments, and technical terminology.

oai_token-act-pair · claude-3-7-sonnet-20250219

Triggered by @neilrathi

The neuron is selectively detecting capitalized names and identifiers (project/script titles) in code comments and headers.

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

Negative Logits

 perifer

-1.34

 awful

-1.29

 бампер

-1.28

ׁ

-1.27

 Medicina

-1.25

llten

-1.24

 充電

-1.21

ली

-1.17

üller

-1.16

không

-1.16

POSITIVE LOGITS

 hvilke

1.30

 mudança

1.23

 duquel

1.23

 hvilken

1.21

 ceea

1.20

 what

1.20

 agujero

1.18

 églises

1.18

 poderá

1.18

 tidligere

1.18

Activations Density 0.006%

Saturday Night Live, Disney, Biolabs

programming or technical content like code headers, script comments, and technical terminology.

The neuron is selectively detecting capitalized names and identifiers (project/script titles) in code comments and headers.

No Known Activations

Saturday Night Live, Disney, Biolabs

programming or technical content like code headers, script comments, and technical terminology.

The neuron is selectively detecting capitalized names and identifiers (project/script titles) in code comments and headers.

No Known Activations