INDEX

Explanations

motions or actions described using the gerund form

oai_token-act-pair · gpt-3.5-turbo

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

©¶æ¥µ

-0.86

Ħ¢

-0.85

HAHA

-0.84

ĪĴ

-0.84

«

-0.82

LEY

-0.79

İĭ

-0.78

¿½

-0.77

Ĥ¬

-0.75

HAEL

-0.75

POSITIVE LOGITS

ast

0.99

actory

0.98

asted

0.95

uzz

0.95

umbers

0.95

ickets

0.94

aunted

0.93

esides

0.93

umped

0.92

ynamic

0.91

Activations Density 0.077%

motions or actions described using the gerund form

No Comments

No Known Activations