INDEX

Explanations

reasoning and logical arguments

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

ãĥİ

-0.81

 Summit

-0.79

omorphic

-0.78

trop

-0.69

atha

-0.67

 Blanc

-0.66

idi

-0.66

EMA

-0.66

 pole

-0.66

 decriminal

-0.66

POSITIVE LOGITS

rists

0.89

you

0.81

 experien

0.81

it

0.79

Ĥ¬

0.73

lamm

0.72

he

0.72

osi

0.71

 they

0.71

 Malk

0.71

Activations Density 14.300%

reasoning and logical arguments

No Comments

No Known Activations