INDEX

Explanations

mentions of political figures

oai_token-act-pair · gpt-3.5-turbo

mentions of political figures, particularly Clinton, Trump, and Obama

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Leilan

-0.64

ulu

-0.60

 Siber

-0.59

 Carbuncle

-0.56

 Constantin

-0.54

 à¨

-0.54

à¦

-0.54

ĪĴ

-0.54

ãĤ¨ãĥ«

-0.53

;;

-0.52

POSITIVE LOGITS

 overflow

0.57

ratulations

0.55

cffff

0.55

jriwal

0.55

 Republicans

0.55

 Democrats

0.53

clinton

0.53

 Bernie

0.52

Sanders

0.52

 hostage

0.52

Activations Density 0.571%

mentions of political figures

mentions of political figures, particularly Clinton, Trump, and Obama

No Comments

No Known Activations