INDEX

Explanations

words related to the possibility of someone being charged with a crime or facing legal consequences

oai_token-act-pair · gpt-3.5-turbo

instances of the verb "be" and its conjugations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 VIDEOS

-0.72

 continuum

-0.66

ciating

-0.65

 promise

-0.64

 specificity

-0.64

eways

-0.64

 Lines

-0.63

 Extend

-0.63

ado

-0.63

 Splash

-0.62

POSITIVE LOGITS

 able

1.30

 punished

1.00

heading

0.98

 subjected

0.97

 forgiven

0.97

 rewarded

0.96

 unable

0.96

 tempted

0.95

 assassinated

0.93

 indicted

0.92

Activations Density 0.280%

words related to the possibility of someone being charged with a crime or facing legal consequences

instances of the verb "be" and its conjugations

No Comments

No Known Activations