INDEX

Explanations

prepositions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "to."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

don

-0.76

 differed

-0.74

irez

-0.73

Written

-0.72

didn

-0.68

ended

-0.67

^^

-0.66

 didn

-0.66

rar

-0.65

puff

-0.64

POSITIVE LOGITS

 downright

0.88

 outright

0.85

 encompass

0.80

pless

0.73

asted

0.73

 adulthood

0.71

DonaldTrump

0.70

wered

0.68

 ensure

0.68

ilet

0.67

Activations Density 0.069%

prepositions

instances of the word "to."

No Comments

No Known Activations