INDEX

Explanations

legal or regulatory terms such as "established," "imposed," "set," "put in place," and "initiated."

oai_token-act-pair · gpt-3.5-turbo

phrases related to regulations and established protocols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

ilater

-0.66

ibli

-0.65

angu

-0.62

oubt

-0.61

ilet

-0.61

pan

-0.59

razil

-0.58

fuck

-0.58

dayName

-0.57

ouf

-0.57

POSITIVE LOGITS

by

1.01

 earlier

0.82

 jointly

0.82

 during

0.80

 previously

0.79

 recently

0.78

 pursuant

0.75

 instituted

0.75

 incidentally

0.73

pez

0.68

Activations Density 0.316%

legal or regulatory terms such as "established," "imposed," "set," "put in place," and "initiated."

phrases related to regulations and established protocols

No Comments

No Known Activations