INDEX

Explanations

references to patronage or support

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "patronage" and associated terms

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

nesota

-0.80

REE

-0.79

MO

-0.73

grass

-0.72

expensive

-0.71

 Matter

-0.70

esm

-0.69

RAM

-0.67

ondo

-0.67

eat

-0.66

POSITIVE LOGITS

 patron

1.00

 patrons

1.00

hetically

0.95

isance

0.86

 saint

0.82

 Patron

0.78

iless

0.75

umeric

0.71

sth

0.71

atern

0.70

Activations Density 0.016%

references to patronage or support

references to the concept of "patronage" and associated terms

No Comments

No Known Activations