INDEX

Explanations

phrases related to governmental activities and politics

oai_token-act-pair · gpt-3.5-turbo

keywords related to governmental or political processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

Defense

-0.77

 Whitman

-0.77

caliber

-0.73

avored

-0.73

 favored

-0.72

rollment

-0.70

colored

-0.67

¥µ

-0.67

imize

-0.66

 endeavor

-0.66

POSITIVE LOGITS

 recognise

1.22

 realise

1.20

 UKIP

1.16

 Â£

1.16

 Labour

1.15

 Ukip

1.15

 apologised

1.14

 centre

1.14

 apologise

1.13

 criticised

1.12

Activations Density 3.796%

phrases related to governmental activities and politics

keywords related to governmental or political processes

No Comments

No Known Activations