INDEX

Explanations

the word "reps" or variations of it

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "reps" and related terms indicating repetition or frequency

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

 Spit

-0.71

ous

-0.70

OUS

-0.68

 Strait

-0.67

 Blow

-0.66

 Galile

-0.64

Gas

-0.64

 tongues

-0.63

 Fare

-0.63

 Crimes

-0.62

POSITIVE LOGITS

 tremend

1.17

è¦ļéĨĴ

1.12

undown

1.10

onse

1.05

utations

0.99

nyder

0.97

arious

0.92

isters

0.91

heet

0.90

arser

0.88

Activations Density 0.006%

the word "reps" or variations of it

occurrences of the word "reps" and related terms indicating repetition or frequency

No Comments

No Known Activations