INDEX

Explanations

proper names like "Nakagawa" and "Hirshfeld"

oai_token-act-pair · gpt-3.5-turbo

mentions of specific names, particularly "Nak" and variations thereof

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

 ILCS

-0.71

DVD

-0.65

 dism

-0.64

 Hercules

-0.62

 alcoholic

-0.62

 Pyro

-0.61

 microwave

-0.61

 Devon

-0.60

 Cook

-0.59

":["

-0.59

POSITIVE LOGITS

amura

1.22

amoto

1.18

Nak

1.15

imura

1.01

omi

0.98

itary

0.98

irin

0.93

aya

0.91

hod

0.90

aji

0.88

Activations Density 0.010%

proper names like "Nakagawa" and "Hirshfeld"

mentions of specific names, particularly "Nak" and variations thereof

No Comments

No Known Activations