INDEX

Explanations

opening parentheses and related characters in code snippets

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

code structure delimiters

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

transQ

-0.54

fillType

-0.50

 ligiloj

-0.49

évaluateur

-0.49

:✨

-0.47

bitat

-0.47

IntoConstraints

-0.47

 queſta

-0.47

تفصیلات

-0.47

ApiModelProperty

-0.46

POSITIVE LOGITS

Nationalité

0.49

EndContext

0.46

0.35

 odkazy

0.35

})(

0.35

 disse

0.34

msleep

0.34

原始内容存档于

0.34

cho

0.34

 Ting

0.34

Activations Density 0.007%

opening parentheses and related characters in code snippets

code structure delimiters

No Comments

No Known Activations