INDEX

Explanations

phrases related to granting permission or approval

oai_token-act-pair · gpt-3.5-turbo

instances of the word "to" and phrases indicating actions or permissions related to it

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.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.6.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

quickShipAvailable

-0.98

furt

-0.74

lies

-0.68

lights

-0.66

grounds

-0.65

mails

-0.65

Reviewed

-0.62

 Needs

-0.61

ELD

-0.60

listed

-0.59

POSITIVE LOGITS

 negotiate

1.11

 indulge

1.09

 avoid

1.08

 achieve

1.05

 pursue

1.04

 gain

1.03

 obtain

1.02

 preserve

1.01

 participate

1.01

 acquire

1.00

Activations Density 0.142%

phrases related to granting permission or approval

instances of the word "to" and phrases indicating actions or permissions related to it

No Comments

No Known Activations