INDEX

Explanations

statements related to government responsibility and potential consequences

oai_token-act-pair · gpt-3.5-turbo

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

colored

-1.10

defense

-1.03

avored

-0.96

 favors

-0.82

rollment

-0.82

Defense

-0.80

caliber

-0.80

imize

-0.80

avior

-0.78

 Shiite

-0.77

POSITIVE LOGITS

 recognise

1.65

 realise

1.61

 recognised

1.50

 apologise

1.40

 Whilst

1.37

 Ukip

1.35

 paed

1.34

 behaviours

1.34

 whilst

1.34

 recogn

1.32

Activations Density 1.792%

statements related to government responsibility and potential consequences

No Comments

No Known Activations