INDEX

Explanations

the word "Palin" with varying activations

oai_token-act-pair · gpt-3.5-turbo

mentions of Sarah Palin

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

Ö¼

-0.83

Increases

-0.74

Trend

-0.74

URI

-0.73

~~~~~~~~~~~~~~~~

-0.73

————————————————

-0.72

upon

-0.72

âĸº

-0.68

Detailed

-0.68

âĵĺ

-0.68

POSITIVE LOGITS

 Palin

1.31

igon

0.97

erity

0.87

istani

0.84

isma

0.83

phrine

0.78

EStream

0.78

iband

0.77

jriwal

0.76

 Schwarzenegger

0.76

Activations Density 0.006%

the word "Palin" with varying activations

mentions of Sarah Palin

No Comments

No Known Activations