INDEX

Explanations

the word "whether" in various contexts

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

whether or conditions

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

 procedure

-0.36

fileID

-0.34

 General

-0.33

and

-0.33

UserDetails

-0.32

PageData

-0.32

Referencias

-0.31

 instruction

-0.31

 dada

-0.31

KeyEvent

-0.31

POSITIVE LOGITS

whether

0.94

 WHETHER

0.86

Whether

0.83

 whether

0.83

 Whether

0.81

 سواء

0.71

MLLoader

0.69

DotNetBar

0.69

 kaarangay

0.68

 jestli

0.63

Activations Density 0.006%

the word "whether" in various contexts

whether or conditions

No Comments

No Known Activations