INDEX

Explanations

locations or institutions and words related to professional qualifications

oai_token-act-pair · gpt-3.5-turbo

references to organizations or studies related to health or medicine

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

Hash

-0.58

 derby

-0.53

ãĤ´ãĥ³

-0.53

amaru

-0.52

 Seym

-0.52

 thous

-0.51

Empty

-0.51

izont

-0.51

eln

-0.50

ãģį

-0.50

POSITIVE LOGITS

 grate

0.65

CVE

0.61

etheless

0.58

 largeDownload

0.57

ciples

0.57

merce

0.55

pin

0.52

soType

0.52

UTERS

0.51

 affirmed

0.50

Activations Density 1.174%

locations or institutions and words related to professional qualifications

references to organizations or studies related to health or medicine

No Comments

No Known Activations