INDEX

Explanations

warnings or things that require careful attention

oai_token-act-pair · gpt-3.5-turbo

terms associated with caution, evaluation, and varying degrees of intensity in experiences and actions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

istor

-0.61

ghazi

-0.60

hift

-0.60

hest

-0.59

 Polic

-0.58

 fold

-0.58

ortium

-0.57

erella

-0.56

 Laurel

-0.56

 comma

-0.55

POSITIVE LOGITS

smanship

0.93

(>

0.91

ptions

0.90

flows

0.90

levels

0.87

vironments

0.87

 awaits

0.84

ourses

0.84

 doses

0.84

 environments

0.83

Activations Density 0.402%

warnings or things that require careful attention

terms associated with caution, evaluation, and varying degrees of intensity in experiences and actions

No Comments

No Known Activations