INDEX

Explanations

Twitter handles

oai_token-act-pair · gpt-3.5-turbo

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

 Bauer

-0.75

ogie

-0.74

 Hawkins

-0.71

 Julie

-0.71

 Kuro

-0.69

Ñı

-0.69

 disapp

-0.69

 Garry

-0.68

POSITIVE LOGITS

1.35

1.34

Tet

1.26

Ts

1.23

TD

1.22

td

1.21

TT

1.21

ts

1.21

TC

1.20

TS

1.17

Activations Density 0.874%

Twitter handles

No Comments

No Known Activations