INDEX

Explanations

phrases related to adherence or opposition to certain rules or standards

oai_token-act-pair · gpt-3.5-turbo

phrases indicating compliance or alignment with policies or regulations

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

 overcame

-0.64

 sacrific

-0.64

 NIGHT

-0.59

 dumped

-0.58

anners

-0.57

 succeed

-0.57

 Jackets

-0.56

 Doodle

-0.56

peed

-0.56

 knocked

-0.56

POSITIVE LOGITS

escap

1.30

humane

1.23

offensive

1.17

versely

1.08

essential

1.08

 stark

1.02

 jeopardy

1.01

 accordance

1.01

lined

1.01

 essence

1.00

Activations Density 0.099%

phrases related to adherence or opposition to certain rules or standards

phrases indicating compliance or alignment with policies or regulations

No Comments

No Known Activations