INDEX

Explanations

terms related to promotions or advancements in career or events

oai_token-act-pair · gpt-3.5-turbo

mentions of promotions in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

 Refuge

-0.70

pH

-0.68

 PRESIDENT

-0.66

itars

-0.65

 sequest

-0.63

 Cage

-0.62

 electroly

-0.62

 cannons

-0.61

 moss

-0.61

 Hayward

-0.61

POSITIVE LOGITS

xual

0.99

 promotion

0.89

stakes

0.83

olicy

0.81

pires

0.76

wcs

0.75

 promoted

0.74

 tremend

0.74

SHIP

0.73

natal

0.72

Activations Density 0.012%

terms related to promotions or advancements in career or events

mentions of promotions in various contexts

No Comments

No Known Activations