INDEX

Explanations

numbers and statistics

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate quantities or measurements

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ãģ¾

-0.51

azard

-0.47

donald

-0.46

appropriately

-0.46

Ø§ÙĦ

-0.45

ioxid

-0.44

idden

-0.44

DCS

-0.44

angering

-0.43

acea

-0.43

POSITIVE LOGITS

 however

0.75

 meanwhile

0.64

 therefore

0.64

 moreover

0.60

 furthermore

0.53

 also

0.52

 though

0.52

 ---------

0.47

 Morty

0.46

 citiz

0.45

Activations Density 4.863%

numbers and statistics

phrases that indicate quantities or measurements

No Comments

No Known Activations