INDEX

Explanations

political terms and phrases

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to the Republican Party and political dynamics involving conservatives

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

[+

-0.80

ĸļ

-0.69

anuts

-0.69

EVA

-0.67

ulkan

-0.67

 Ranking

-0.66

sels

-0.66

 Miko

-0.66

Ù

-0.65

 Peng

-0.65

POSITIVE LOGITS

 insurgency

1.16

 movement

1.14

 insurgent

1.14

 electorate

1.11

 populism

1.09

 populist

1.08

 grassroots

1.02

Tea

1.02

 revolt

1.01

 insurg

0.99

Activations Density 0.320%

political terms and phrases

references to the Republican Party and political dynamics involving conservatives

No Comments

No Known Activations