INDEX

Explanations

the word "sleeping" and related terms

oai_token-act-pair · gpt-3.5-turbo

references to sleeping or sleep-related activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 certs

-0.79

iferation

-0.75

KC

-0.73

 EDITION

-0.70

Jar

-0.69

credit

-0.65

bury

-0.64

 excise

-0.64

 Newman

-0.64

 Grassley

-0.62

POSITIVE LOGITS

 sleeping

3.61

 Sleeping

2.78

 sleeps

2.46

 slept

2.42

 sleep

2.18

 asleep

1.99

sleep

1.79

Sleep

1.78

 Sleep

1.68

sle

1.67

Activations Density 0.012%

the word "sleeping" and related terms

references to sleeping or sleep-related activities

No Comments

No Known Activations