INDEX

Explanations

keywords related to political news and legislative actions

oai_token-act-pair · gpt-3.5-turbo

references to congressional actions and legislation

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

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Afee

-0.91

 recru

-0.79

 handlers

-0.73

 IMAGES

-0.71

osate

-0.66

 competitor

-0.66

Company

-0.65

VPN

-0.64

Atl

-0.64

itbart

-0.63

POSITIVE LOGITS

 adjourn

1.52

 unanimously

1.41

 recess

1.32

 convened

1.23

 votes

1.19

 deliberations

1.13

 deliber

1.10

 unanimous

1.10

 vote

1.10

 voted

1.10

Activations Density 0.293%

keywords related to political news and legislative actions

references to congressional actions and legislation

No Comments

No Known Activations