INDEX

Explanations

text related to computer system information and technical errors

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

ulhu

-0.82

ificial

-0.81

zona

-0.79

 ACTIONS

-0.77

igham

-0.76

ccording

-0.75

 narrower

-0.73

 Viper

-0.69

othy

-0.69

afety

-0.69

POSITIVE LOGITS

enhagen

0.84

redo

0.83

eties

0.78

acht

0.75

hirt

0.74

camp

0.73

>(

0.72

ppelin

0.71

 counsel

0.69

iuses

0.68

Activations Density 18.896%

text related to computer system information and technical errors

No Comments

No Known Activations