INDEX

Explanations

terms related to dishonesty or double standards

oai_token-act-pair · gpt-3.5-turbo

terms related to hypocrisy and disingenuous behavior

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

league

-0.78

ivan

-0.69

gan

-0.69

Grant

-0.69

gran

-0.68

debian

-0.68

anguages

-0.67

 Sunder

-0.67

kee

-0.66

Trust

-0.66

POSITIVE LOGITS

 hypoc

1.22

 hypocrisy

1.22

 hypocr

1.15

 hypocritical

1.12

 crus

0.83

isSpecialOrderable

0.83

cyn

0.82

 Bastard

0.75

 indignation

0.75

 disgu

0.75

Activations Density 0.012%

terms related to dishonesty or double standards

terms related to hypocrisy and disingenuous behavior

No Comments

No Known Activations