INDEX

Explanations

words related to decisions and actions

oai_token-act-pair · gpt-3.5-turbo

phrases and references to nostalgia and past experiences

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

ãĥīãĥ©

-0.60

gang

-0.56

ammy

-0.52

rocal

-0.51

ideshow

-0.50

 Moines

-0.50

pires

-0.50

ys

-0.49

iple

-0.48

mx

-0.48

POSITIVE LOGITS

 namely

1.25

especially

1.07

 hence

0.98

 albeit

0.93

particularly

0.93

 culminating

0.93

 especially

0.93

 whereas

0.90

 Especially

0.89

 although

0.88

Activations Density 0.537%

words related to decisions and actions

phrases and references to nostalgia and past experiences

No Comments

No Known Activations