INDEX

Explanations

phrases related to governmental or regulatory bodies

oai_token-act-pair · gpt-3.5-turbo

mentions of "Authority" or "Authorities."

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ritz

-0.70

yond

-0.65

flows

-0.65

 streaming

-0.64

 chat

-0.64

pieces

-0.64

ets

-0.63

 stitch

-0.61

 streamed

-0.61

 flowers

-0.60

POSITIVE LOGITS

 Authority

4.18

 authority

2.07

 Authorities

1.84

 Agency

1.43

 AUTHOR

1.43

 Administration

1.17

 authorities

1.17

 Commission

1.05

 Commissioner

1.02

 Advisory

1.01

Activations Density 0.019%

phrases related to governmental or regulatory bodies

mentions of "Authority" or "Authorities."

No Comments

No Known Activations