INDEX

Explanations

prepositions and conjunctions indicating relationships between actions and subjects

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 13-gemmascope-res-65k

Configuration

google/gemma-scope-2b-pt-res/layer_13/width_65k/average_l0_75

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.13.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

and

-0.92

bootstrapcdn

-0.85

évaluateur

-0.63

性和

-0.59

☸

-0.54

 CURIAM

-0.52

хьтан

-0.51

SourceChecksum

-0.51

 Roskov

-0.50

 eivät

-0.49

POSITIVE LOGITS

 abgerufen

0.69

 accordingly

0.66

دانشنامهٔ

0.64

CRIPT

0.62

 noqa

0.60

hyrchwyd

0.60

 дописавши

0.57

AZIONI

0.57

toplasm

0.57

PLICATION

0.53

Activations Density 0.076%

prepositions and conjunctions indicating relationships between actions and subjects

No Comments

No Known Activations