INDEX

Explanations

words related to death and burial

oai_token-act-pair · gpt-3.5-turbo

references to coffins and related imagery

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 Yellowstone

-0.80

 Rogers

-0.75

anni

-0.73

 Trout

-0.66

Å«

-0.66

 hops

-0.65

 redirect

-0.65

 adjud

-0.64

 spotting

-0.64

recogn

-0.64

POSITIVE LOGITS

 coffin

3.99

 coff

2.20

 Coffin

1.63

 Coff

1.39

asket

1.26

 mummy

1.14

 corpse

1.05

icol

0.99

 sarc

0.98

wra

0.97

Activations Density 0.038%

words related to death and burial

references to coffins and related imagery

No Comments

No Known Activations