INDEX

Explanations

events, experiments, or entities that are divided or split into different parts or categories

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

tor

-0.82

enegger

-0.75

die

-0.74

tun

-0.74

enhagen

-0.73

 onwards

-0.72

 challeng

-0.72

 chin

-0.71

WT

-0.70

onda

-0.70

POSITIVE LOGITS

 thirds

1.01

qqa

0.87

 categories

0.85

ãĤ©

0.82

uild

0.81

clusions

0.79

perse

0.78

units

0.77

Æ

0.74

 Tradable

0.72

Activations Density 6.155%

events, experiments, or entities that are divided or split into different parts or categories

No Comments

No Known Activations