INDEX

Explanations

numerical figures or statistics

oai_token-act-pair · gpt-3.5-turbo

references to statistical data or numerical 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

ModLoader

-0.96

GBT

-0.79

vid

-0.78

Interstitial

-0.72

ntil

-0.71

IFE

-0.71

struction

-0.70

aughs

-0.70

ecast

-0.67

umenthal

-0.66

POSITIVE LOGITS

 figures

0.95

heet

0.80

 estimates

0.79

 figure

0.78

enance

0.75

 prominently

0.72

oil

0.72

 charts

0.72

 compiled

0.71

hig

0.71

Activations Density 0.024%

numerical figures or statistics

references to statistical data or numerical information

No Comments

No Known Activations