INDEX

Explanations

imperative statements or instructions given to someone

oai_token-act-pair · gpt-3.5-turbo

commands or directives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

quickShipAvailable

-0.82

 Appears

-0.78

grounds

-0.78

pires

-0.75

MpServer

-0.73

cano

-0.72

DAQ

-0.70

 largeDownload

-0.70

finder

-0.67

imilar

-0.65

POSITIVE LOGITS

 keep

0.94

 evacuate

0.94

 refrain

0.92

 protect

0.92

 stop

0.91

 shove

0.91

 beware

0.90

 treat

0.90

 avoid

0.90

 drink

0.89

Activations Density 0.135%

imperative statements or instructions given to someone

commands or directives

No Comments

No Known Activations