INDEX

Explanations

descriptions or mentions of lids

oai_token-act-pair · gpt-3.5-turbo

references to lids and their significance or usage

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Capital

-0.75

present

-0.71

bidden

-0.71

fortune

-0.69

capital

-0.68

WER

-0.67

 Blessed

-0.66

gotten

-0.65

represented

-0.65

 Tribe

-0.62

POSITIVE LOGITS

lid

1.45

ysis

1.07

wcs

0.80

 burner

0.80

creen

0.80

plates

0.80

 curtain

0.79

bell

0.79

heed

0.78

 curtains

0.77

Activations Density 0.006%

descriptions or mentions of lids

references to lids and their significance or usage

No Comments

No Known Activations