INDEX

Explanations

elements related to image processing and positional data structures

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

dimensions and coordinates

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-131k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_131k/average_l0_109

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

131,072

Data Type

float32

Hook Name

blocks.31.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

ArrowToggle

-0.56

AutoScaleMode

-0.52

/**

-0.48

EndContext

-0.47

 ExecuteAsync

-0.44

/*

-0.43

lankton

-0.43



-0.42

 calldata

-0.41

 Verſ

-0.41

POSITIVE LOGITS

 latitude

0.53

latitude

0.50

Rows

0.46

0.45

"("

0.44

width

0.43

 rows

0.43

rows

0.42

 height

0.41

 width

0.41

Activations Density 0.093%

elements related to image processing and positional data structures

dimensions and coordinates

No Comments

No Known Activations