INDEX

Explanations

names and terms related to individuals or characters

oai_token-act-pair · gpt-3.5-turbo

mentions of specific individuals, particularly those with the name "Kaw"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Establishment

-0.69

xual

-0.69

eering

-0.68

llor

-0.65

 Scotia

-0.63

psy

-0.62

 purge

-0.62

balance

-0.61

istically

-0.61

 Murdoch

-0.60

POSITIVE LOGITS

asaki

1.30

Kaw

1.18

aii

1.15

amoto

1.04

ata

0.93

arij

0.92

oji

0.92

©¶æ¥µ

0.92

atsu

0.90

ota

0.87

Activations Density 0.008%

names and terms related to individuals or characters

mentions of specific individuals, particularly those with the name "Kaw"

No Comments

No Known Activations