INDEX

Explanations

political figures and terms

oai_token-act-pair · gpt-3.5-turbo

references to political party leadership roles, specifically in the context of legislative discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

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

atars

-0.79

worldly

-0.70

 aval

-0.65

nib

-0.65

 caliphate

-0.65

Sov

-0.64

ãĤ¡

-0.64

estamp

-0.64

 needles

-0.63

nai

-0.63

POSITIVE LOGITS

 Leader

1.76

 Whip

1.61

Leader

1.34

 Majority

1.11

 Caucus

1.07

 leader

1.05

 Leaders

1.03

 Floor

1.02

 Minority

0.99

 Speaker

0.95

Activations Density 0.038%

political figures and terms

references to political party leadership roles, specifically in the context of legislative discussions

No Comments

No Known Activations