INDEX

Explanations

references to boundaries and related political terms

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

references to geographic boundaries or divisions

oai_token-act-pair · gpt-3.5-turbo

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

hiba

-0.90

eah

-0.89

ems

-0.88

orah

-0.87

ety

-0.82

cffffcc

-0.82

efully

-0.81

etooth

-0.80

heed

-0.80

ebin

-0.78

POSITIVE LOGITS

 Kane

0.90

VO

0.66

 Allan

0.66

 Lauder

0.65

 Hebdo

0.65

 Ware

0.64

 Dame

0.63

 Clear

0.63

naire

0.63

 Wast

0.62

Activations Density 0.123%

references to boundaries and related political terms

references to geographic boundaries or divisions

No Comments

No Known Activations