INDEX

Explanations

a broad array of generally short and common words, and appears to assign higher relevance when multiple of these words appear in sequence.

oai_token-act-pair · gemini-2.0-flash

Clear/clean

np_max-act-logits · gemini-2.0-flash

New Auto-Interp

Configuration

google/gemma-scope-2b-pt-transcoders/layer_25/width_16k/average_l0_41

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.25.ln2.hook_normalized

Architecture

jumprelu_transcoder

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 sureties

-0.60

География

-0.58

inese

-0.52

حياتها

-0.50

djang

-0.48

 olsun

-0.47

toMatchSnapshot

-0.47

 чек

-0.47

stdafx

-0.46

riwal

-0.46

POSITIVE LOGITS

 clear

2.30

clear

2.02

 clean

1.91

 Clear

1.80

Clear

1.74

clean

1.59

 CLEAR

1.42

 Clean

1.41

Clean

1.37

CLEAR

1.30

Activations Density 5.786%

a broad array of generally short and common words, and appears to assign higher relevance when multiple of these words appear in sequence.

Clear/clean

No Comments

No Known Activations

a broad array of generally short and common words, and appears to assign higher relevance when multiple of these words appear in sequence.

Clear/clean

No Comments

No Known Activations