INDEX

Explanations

phrases related to key findings or conclusions

oai_token-act-pair · gpt-3.5-turbo

key points or conclusions from discussions or reports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

tiny

-0.74

nder

-0.72

ements

-0.72

iop

-0.71

ently

-0.71

cking

-0.70

umbered

-0.69

ifling

-0.67

 slightest

-0.67

este

-0.67

POSITIVE LOGITS

 takeaway

0.99

PsyNetMessage

0.79

DragonMagazine

0.79

actionDate

0.75

ODUCT

0.71

é¾įå

0.71

 glean

0.70

é»Ĵ

0.69

 Opportun

0.68

 Learned

0.68

Activations Density 0.018%

phrases related to key findings or conclusions

key points or conclusions from discussions or reports

No Comments

No Known Activations