INDEX

Explanations

phrases related to gathering information or gaining insight

oai_token-act-pair · gpt-3.5-turbo

phrases and concepts related to clarity and understanding in information

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

bees

-0.65

ammy

-0.59

ovie

-0.58

ueller

-0.56

hate

-0.56

cessive

-0.55

cakes

-0.53

 expel

-0.52

 champagne

-0.52

 anth

-0.52

POSITIVE LOGITS

DragonMagazine

0.82

 detailing

0.80

PsyNetMessage

0.79

 regarding

0.75

Discuss

0.72

 opin

0.69

ãĤ¶

0.66

wcs

0.66

phabet

0.66

éĸ

0.65

Activations Density 0.239%

phrases related to gathering information or gaining insight

phrases and concepts related to clarity and understanding in information

No Comments

No Known Activations