INDEX

Explanations

phrases related to decisions or statements made by individuals or organizations

oai_token-act-pair · gpt-3.5-turbo

references to decisions, rulings, and announcements in a formal context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 behav

-0.63

 liking

-0.60

 dying

-0.60

Joined

-0.59

 hobbies

-0.58

 hemor

-0.58

ghost

-0.57

 envy

-0.56

Unix

-0.56

training

-0.55

POSITIVE LOGITS

 underscores

1.13

 signifies

1.07

 reinforces

1.03

 represents

1.03

 proves

1.00

 illustrates

0.99

 demonstrates

0.99

 reminds

0.98

 reflects

0.95

 brings

0.93

Activations Density 0.484%

phrases related to decisions or statements made by individuals or organizations

references to decisions, rulings, and announcements in a formal context

No Comments

No Known Activations