INDEX

Explanations

phrases related to products or services being new or recently introduced

oai_token-act-pair · gpt-3.5-turbo

the phrase "brand new" and its various occurrences

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

absor

-0.72

prov

-0.71

eme

-0.67

ocratic

-0.63

erves

-0.62

Ĥ

-0.61

upper

-0.61

benef

-0.61

Ju

-0.61

zik

-0.61

POSITIVE LOGITS

ciating

0.73

bies

0.70

 edition

0.68

 Frankenstein

0.66

 storefront

0.66

adoes

0.65

ilings

0.65

ismo

0.64

 iceberg

0.64

 scenery

0.64

Activations Density 0.044%

phrases related to products or services being new or recently introduced

the phrase "brand new" and its various occurrences

No Comments

No Known Activations