INDEX

Explanations

patterns or structures in coding or programming contexts

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

code keywords and abbreviations

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

if

-0.42

 also

-0.40

do

-0.36

if

-0.35

and

-0.35

Also

-0.34

en

-0.34

POSITIVE LOGITS

 ModelExpression

1.03

OGND

0.93

 الحره

0.86

✨:

0.84

twimg

0.82

endphp

0.81

<unused8>

0.80

 للاسماء

0.80

[@BOS@]

0.79

<unused43>

0.79

Activations Density 0.006%

patterns or structures in coding or programming contexts

code keywords and abbreviations

No Comments

No Known Activations