INDEX

Explanations

phrases indicating a comparison or measurement of an outcome or result

oai_token-act-pair · gpt-3.5-turbo

phrases that express the concept of something "amounting to" a certain level or significance

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

tun

-0.79

 craw

-0.79

notations

-0.71

notation

-0.70

 continues

-0.65

reys

-0.63

builders

-0.62

rers

-0.62

uve

-0.60

 reminded

-0.59

POSITIVE LOGITS

anuts

0.72

 perfection

0.70

 superhuman

0.66

oteric

0.64

aea

0.63

 heresy

0.62

 save

0.61

 something

0.60

an

0.60

Interstitial

0.59

Activations Density 0.075%

phrases indicating a comparison or measurement of an outcome or result

phrases that express the concept of something "amounting to" a certain level or significance

No Comments

No Known Activations