INDEX

Explanations

programming language syntax structures

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

conditional statements (if/else)

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

exitRule

-0.73

+#+

-0.59

 NSCoder

-0.56

bootstrapcdn

-0.54

delwed

-0.54

fxml

-0.52

хьтан

-0.51

InSection

-0.51

 تضيفلها

-0.50

 CanadaChoose

-0.49

POSITIVE LOGITS

 للاسماء

0.41

 wypad

0.37

 whether

0.34

Вывод

0.32

try

0.32

 canzone

0.31

 hypothesis

0.31

ってみて

0.31

 vinden

0.31

 calculer

0.31

Activations Density 0.009%

programming language syntax structures

conditional statements (if/else)

No Comments

No Known Activations