INDEX

Explanations

locations associated with military bases or prisons

oai_token-act-pair · gpt-3.5-turbo

references to specific locations and individuals associated with political or military contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

-+-+-+-+

-0.79

ãĤ©

-0.79

rar

-0.76

ibilities

-0.76

ãĥ¼ãĤ¯

-0.75

Reviewer

-0.70

Ðº

-0.67

ãĥ¼ãĥ

-0.65

ãĥĥ

-0.65

eline

-0.65

POSITIVE LOGITS

 Foundation

0.70

Jr

0.66

 Gold

0.66

 Unlimited

0.65

enium

0.65

 Slim

0.65

 Quart

0.64

Sr

0.64

Ltd

0.63

XX

0.62

Activations Density 0.220%

locations associated with military bases or prisons

references to specific locations and individuals associated with political or military contexts

No Comments

No Known Activations