INDEX

Explanations

references to unexpected or surprising plot developments, often described as "twists."

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

league

-0.65

leased

-0.61

ufact

-0.60

hammad

-0.57

 Ducks

-0.57

RES

-0.55

ogl

-0.55

rahim

-0.54

 appropriation

-0.54

usable

-0.53

POSITIVE LOGITS

 twist

1.07

 twists

0.99

 Twist

0.81

 endings

0.71

 twisting

0.71

 Whedon

0.65

 helic

0.64

 angle

0.64

 weave

0.64

endish

0.63

Activations Density 10.693%

references to unexpected or surprising plot developments, often described as "twists."

No Comments

No Known Activations