INDEX

Explanations

references to editors in various publications

oai_token-act-pair · gpt-3.5-turbo

references to editors in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

bley

-0.78

pered

-0.78

bones

-0.74

enegger

-0.74

pering

-0.70

ulia

-0.69

achi

-0.69

ptoms

-0.69

ffff

-0.68

ulus

-0.66

POSITIVE LOGITS

ials

1.09

ially

1.02

furt

0.85

editor

0.83

 Picks

0.74

ienne

0.69

 editor

0.68

 Zack

0.68

 laureate

0.68

 Trevor

0.66

Activations Density 0.015%

references to editors in various publications

references to editors in various contexts

No Comments

No Known Activations