INDEX

Explanations

words associated with legitimacy or legal validity

oai_token-act-pair · gpt-3.5-turbo

terms related to legitimacy and authenticity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

glass

-0.78

gt

-0.73

pan

-0.68

 compares

-0.64

lin

-0.64

 Cats

-0.63

 Compass

-0.62

lein

-0.62

 Slip

-0.61

ank

-0.61

POSITIVE LOGITS

 legitimate

3.56

 legit

2.35

 legitimately

2.27

 genuine

1.85

itimate

1.83

 illegitimate

1.81

 valid

1.77

 rightful

1.77

 lawful

1.65

 bona

1.57

Activations Density 0.014%

words associated with legitimacy or legal validity

terms related to legitimacy and authenticity

No Comments

No Known Activations