INDEX

Explanations

text with a high degree of ellipses and placeholders

oai_token-act-pair · gpt-4o-mini Triggered by @bot

.

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.20.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Alger

-0.44

ftagPool

-0.44

fortawesome

-0.43

giarism

-0.42

FunctionFlags

-0.42

MERCIAL

-0.42

"])

-0.41

ized

-0.41

 Administrativna

-0.41

ctez

-0.41

POSITIVE LOGITS

,...

0.71

...

0.69

...

0.64

?...

0.63

'..

0.63

/...

0.62

 ....

0.61

,..

0.61

..

0.61

:...

0.61

Activations Density 0.036%

text with a high degree of ellipses and placeholders

.

No Comments

No Known Activations