INDEX

Explanations

mathematical symbols and expressions

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

IsMutable

-0.71

-0.58

 wireType

-0.58

 <<<<<<<<<<<<<<

-0.58

 cherchés

-0.57

__*/

-0.56

 ſei

-0.56

Havolalar

-0.55

 poffible

-0.55

 beſte

-0.54

POSITIVE LOGITS

 zweiten

0.43

 ikinci

0.42

second

0.42

 second

0.40

 tweede

0.40

 secondary

0.40

 secundaria

0.40

Secondary

0.40

 zweite

0.40

secondary

0.39

Activations Density 0.433%

mathematical symbols and expressions

No Comments

No Known Activations