INDEX

Explanations

dates or time periods

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing historical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wcsstore

-0.94

abul

-0.84

*/(

-0.83

ittees

-0.79

rouse

-0.77

erity

-0.77

ptoms

-0.76

earchers

-0.76

umerable

-0.76

ħĭ

-0.72

POSITIVE LOGITS

 mankind

1.35

 humankind

1.32

 warfare

1.11

 humanity

1.06

 civilization

1.04

 civilisation

0.96

 colonialism

0.92

 Computing

0.91

 Christianity

0.91

 medicine

0.90

Activations Density 0.147%

dates or time periods

phrases emphasizing historical contexts

No Comments

No Known Activations