INDEX

Explanations

social media related terms and phrases

oai_token-act-pair · gpt-3.5-turbo

mentions of the messaging platform WhatsApp

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

Closure

-0.68

cean

-0.63

fitting

-0.61

bri

-0.61

Gou

-0.61

 demolition

-0.60

 Ramsey

-0.60

iling

-0.59

REF

-0.59

JM

-0.59

POSITIVE LOGITS

creen

0.93

bour

0.87

ocial

0.87

hower

0.87

rael

0.83

omething

0.83

iques

0.82

ilon

0.81

cha

0.80

ername

0.78

Activations Density 0.017%

social media related terms and phrases

mentions of the messaging platform WhatsApp

No Comments

No Known Activations