INDEX

Explanations

phrases related to personal beliefs and political views

oai_token-act-pair · gpt-3.5-turbo

references to political and personal beliefs

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

eri

-0.70

Luck

-0.67

 Craw

-0.65

amaz

-0.65

owl

-0.63

enary

-0.62

pletion

-0.61

 Hitch

-0.61

Kn

-0.61

 unbeliev

-0.61

POSITIVE LOGITS

 about

1.02

 expressed

0.92

esp

0.90

 concerning

0.88

 beliefs

0.88

 stances

0.82

 regarding

0.82

 clashed

0.81

 ABOUT

0.78

 align

0.75

Activations Density 0.177%

phrases related to personal beliefs and political views

references to political and personal beliefs

No Comments

No Known Activations