INDEX

Explanations

proper nouns

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ptions

-0.83

fecture

-0.83

eer

-0.81

places

-0.72

ebus

-0.72

orers

-0.71

spring

-0.71

sites

-0.71

leased

-0.69

yrinth

-0.67

POSITIVE LOGITS

 Jefferson

1.00

Aqu

0.97

 Mulcair

0.93

 Browne

0.89

 Hardy

0.83

 Edison

0.82

 Crom

0.82

 Carly

0.74

aeus

0.73

 Beard

0.71

Activations Density 0.020%

proper nouns

No Comments

No Known Activations