INDEX

Explanations

verbs related to initiating or beginning actions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "start" and its variations 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 @ 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

obi

-0.75

ighth

-0.74

 entirety

-0.69

cedented

-0.68

odes

-0.68

itsch

-0.68

alted

-0.66

abilities

-0.66

acho

-0.66

ugs

-0.65

POSITIVE LOGITS

 anew

0.99

 behaving

0.79

nings

0.76

starting

0.74

 PRESS

0.73

 raining

0.73

 wondering

0.71

 fuss

0.70

strap

0.69

 screaming

0.68

Activations Density 0.071%

verbs related to initiating or beginning actions

instances of the word "start" and its variations in various contexts

No Comments

No Known Activations