INDEX

Explanations

locations related to theaters and performance spaces

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

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

RemoveField

-0.40

 mental

-0.33

 blow

-0.32

 inter

-0.32

Manbalar

-0.31

 motivated

-0.31

 mortal

-0.30

ொ

-0.30

nek

-0.30

 flavor

-0.30

POSITIVE LOGITS

 auditorium

0.68

 theatres

0.68

 theatre

0.68

 theater

0.65

theatre

0.65

 Theater

0.65

 Theatre

0.64

theater

0.64

SBATCH

0.63

 theaters

0.61

Activations Density 0.011%

locations related to theaters and performance spaces

No Comments

No Known Activations