INDEX

Explanations

references to literary works, specifically novels

oai_token-act-pair · gpt-3.5-turbo

references to novels and literature

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

xon

-0.86

Downloadha

-0.71

 impunity

-0.69

rael

-0.65

repair

-0.65

olid

-0.62

older

-0.62

aples

-0.61

âĹ¼

-0.60

areth

-0.60

POSITIVE LOGITS

ties

1.43

izations

1.32

isations

1.17

ization

1.16

isation

1.16

istic

1.04

 manuscript

0.99

istically

0.99

istics

0.94

ists

0.92

Activations Density 0.031%

references to literary works, specifically novels

references to novels and literature

No Comments

No Known Activations