INDEX

Explanations

phrases related to ongoing debates

oai_token-act-pair · gpt-3.5-turbo

instances of the word "debate" and related discussions surrounding various topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

oho

-0.67

fits

-0.66

iaries

-0.65

amina

-0.65

urance

-0.63

urances

-0.63

istically

-0.59

photos

-0.58

 liner

-0.58

uras

-0.58

POSITIVE LOGITS

 raging

1.00

 raged

1.00

 debate

0.86

 debates

0.85

 regarding

0.85

 concerning

0.85

 debated

0.84

halla

0.83

naire

0.79

 debating

0.77

Activations Density 0.061%

phrases related to ongoing debates

instances of the word "debate" and related discussions surrounding various topics

No Comments

No Known Activations