INDEX

Explanations

technical instructions or steps in a process

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

BuyableInstoreAndOnline

-0.66

 pestic

-0.64

VIDIA

-0.62

putable

-0.61

ITNESS

-0.61

NetMessage

-0.59

 persuasion

-0.58

 consolation

-0.58

 confir

-0.56

Maw

-0.55

POSITIVE LOGITS

 accordingly

1.06

into

0.86

 into

0.85

 onto

0.78

ifice

0.75

til

0.72

 appropriately

0.69

onto

0.69

selves

0.69

 unnecessarily

0.67

Activations Density 20.864%

technical instructions or steps in a process

No Comments

No Known Activations