INDEX

Explanations

references to style and design elements in code or descriptions

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

UI styling properties

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

AddTagHelper

-0.64

writeFieldEnd

-0.58

awaiter

-0.56

TintMode

-0.52

 poffe

-0.51

Tikang

-0.49

AsUp

-0.49

 raiſ

-0.48

 Jefus

-0.48

ngrx

-0.47

POSITIVE LOGITS

 StyleSheet

0.61

 useStyles

0.55

styles

0.52

IContainer

0.47

 initComponents

0.45

 makeStyles

0.43

MigrationBuilder

0.41

lehrer

0.39

Styles

0.37

 dignité

0.36

Activations Density 0.000%

references to style and design elements in code or descriptions

UI styling properties

No Comments

No Known Activations