INDEX

Explanations

mathematical terms and formulas

oai_token-act-pair · gpt-3.5-turbo

terms related to mathematical concepts and notation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

agers

-0.74

 hire

-0.72

 womb

-0.68

 sher

-0.63

chnology

-0.63

saf

-0.63

 Hond

-0.62

 pressing

-0.62

vag

-0.60

 combust

-0.60

POSITIVE LOGITS

math

1.43

DragonMagazine

0.98

ãĤ¸

0.96

ãĤ¶

0.92

TeX

0.83

ãĤ¡

0.82

è£ħ

0.82

 fixme

0.80

ãĥ´ãĤ¡

0.80

ãĤ·

0.79

Activations Density 0.007%

mathematical terms and formulas

terms related to mathematical concepts and notation

No Comments

No Known Activations