INDEX

Explanations

phrases or words that are critical or judgmental in nature

oai_token-act-pair · gpt-3.5-turbo

instances of remarks or statements that are noteworthy or significant

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

ullivan

-0.74

 funnel

-0.72

 tunnel

-0.68

ccording

-0.66

srf

-0.64

 conflic

-0.64

ãĥ«

-0.63

wired

-0.61

ramid

-0.61

nels

-0.61

POSITIVE LOGITS

eting

0.94

ments

0.85

iasis

0.84

ful

0.83

men

0.80

gow

0.80

ries

0.80

ciation

0.79

brate

0.78

erence

0.77

Activations Density 0.048%

phrases or words that are critical or judgmental in nature

instances of remarks or statements that are noteworthy or significant

No Comments

No Known Activations