INDEX

Explanations

references to repetition, additional elements, and layers in a process or description

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

BitmapFactory

-0.37

 dabei

-0.35

 jahr

-0.34

 bäst

-0.33

lepiej

-0.30

background

-0.29

 background

-0.29

 gern

-0.29

 pues

-0.28

äumt

-0.28

POSITIVE LOGITS

ConstraintMaker

0.71

 ब्रेकडाउन

0.67

 kaarangay

0.67

WebElementEntity

0.66

IntoConstraints

0.62

AntiForgeryToken

0.60

yntaxException

0.59

SBATCH

0.59

bleau

0.58

AlterField

0.57

Activations Density 0.095%

references to repetition, additional elements, and layers in a process or description

No Comments

No Known Activations