INDEX

Explanations

proper names related to individuals

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Lar" and variations of it

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

ãģ®éŃĶ

-0.98

ä»£

-0.86

éĹĺ

-0.83

manship

-0.81

 Marketable

-0.76

ħĭ

-0.75

å§«

-0.75

CRIP

-0.73

å£«

-0.73

ãģĭ

-0.72

POSITIVE LOGITS

Lar

1.15

acan

0.98

oche

0.95

issa

0.92

sson

0.91

vae

0.90

acy

0.87

imore

0.86

abee

0.85

ocious

0.84

Activations Density 0.006%

proper names related to individuals

occurrences of the name "Lar" and variations of it

No Comments

No Known Activations