INDEX

Explanations

details related to software, data, and code development

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

orc

-0.92

aminer

-0.84

ensional

-0.81

arling

-0.75

anguage

-0.74

ellation

-0.73

othy

-0.73

ensed

-0.72

ente

-0.71

agos

-0.70

POSITIVE LOGITS

th

0.88

rd

0.75

âĸĪâĸĪ

0.71

eteen

0.65

 cents

0.64

 stitches

0.64

0.63

 jars

0.60

teenth

0.60

Activations Density 7.082%

details related to software, data, and code development

No Comments

No Known Activations