INDEX

Explanations

references to the term "Patriot"

oai_token-act-pair · gpt-3.5-turbo

references to the "Patriot Act" and related terminology

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ð

-0.83

redits

-0.80

flows

-0.79

kell

-0.78

err

-0.78

sl

-0.76

eq

-0.75

step

-0.75

hov

-0.75

forces

-0.75

POSITIVE LOGITS

 Patriot

1.22

riot

1.09

 Scotia

0.91

riots

0.84

 Prayer

0.78

 Seym

0.77

ãĥ¼ãĥĨ

0.74

 Maker

0.74

cia

0.73

 Anthem

0.72

Activations Density 0.014%

references to the term "Patriot"

references to the "Patriot Act" and related terminology

No Comments

No Known Activations