INDEX

Explanations

code snippets and programming language terms

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

code followed by colon

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

-0.74

WireFormatLite

-0.64

ScopeManager

-0.62

 BoxDecoration

-0.62

 transfieras

-0.56

tagHelperRunner

-0.56

 AssemblyCulture

-0.56

 BoxFit

-0.56

########.

-0.55

ViewImports

-0.52

POSITIVE LOGITS

łach

0.34

 виправивши

0.33

imanapun

0.31

 presente

0.31

ocardio

0.29

 fascinating

0.28

 presentes

0.28

 करती

0.27

 Kerk

0.27

Garbage

0.27

Activations Density 0.565%

code snippets and programming language terms

code followed by colon

No Comments

No Known Activations