INDEX

Explanations

specific symbols and characters in a particular context that seem to form a pattern or code, possibly related to culture or language

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Configuration

neuronpedia/gpt2-small__res_scl-ajt/6-res_scl-ajt

Prompts (Dashboard)

12,288 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

46,080

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

apollo-research/Skylion007-openwebtext-tokenizer-gpt2

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

etition

-1.01

ebus

-0.95

awei

-0.89

haw

-0.88

horn

-0.85

sonian

-0.84

itive

-0.82

pick

-0.80

kefeller

-0.79

arton

-0.79

POSITIVE LOGITS

âĶĢâĶĢ

0.85

 Karin

0.77

BuyableInstoreAndOnline

0.77

ãĥĸ

0.76

âĿ

0.76

 Abram

0.75

;;;;;;;;;;;;

0.72

lihood

0.71

 Isles

0.68

INE

0.68

Activations Density 7.698%

specific symbols and characters in a particular context that seem to form a pattern or code, possibly related to culture or language

No Comments

No Known Activations