INDEX

Explanations

references to vacations

oai_token-act-pair · gpt-3.5-turbo

mentions of vacations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ioch

-0.71

 circumst

-0.70

efficients

-0.67

corn

-0.66

rencies

-0.63

-----------

-0.62

ourke

-0.62

ifice

-0.62

ussen

-0.61

uscript

-0.61

POSITIVE LOGITS

 vacation

1.11

 vacations

1.09

Vac

0.89

 rentals

0.86

oleon

0.79

 touring

0.76

 resorts

0.71

 abroad

0.67

spection

0.66

agascar

0.66

Activations Density 0.008%

references to vacations

mentions of vacations

No Comments

No Known Activations