INDEX

Explanations

non-explicit or non-verbal textual elements in programming or markup languages, particularly those involving structures or placeholders

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

New Auto-Interp

Configuration

google/gemma-scope-9b-pt-res/layer_20/width_131k/average_l0_114

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

 vacío

-0.47

atshirt

-0.44

 Segurança

-0.43

 vôtre

-0.43

pantalón

-0.42

kosť

-0.41

 Evangelio

-0.41

 useAppContext

-0.41

 saco

-0.40

 écrivain

-0.40

POSITIVE LOGITS

awtextra

0.71

twimg

0.48

 Gladys

0.44

0.43

THON

0.43

 SwiftUI

0.42

hoeddwyd

0.42

 Pingback

0.41

glor

0.41

 Misty

0.40

Activations Density 0.000%

non-explicit or non-verbal textual elements in programming or markup languages, particularly those involving structures or placeholders

No Comments

No Known Activations