INDEX

Explanations

proper nouns related to a specific name, "Kotick"

oai_token-act-pair · gpt-3.5-turbo

references to the name "Kot" in various contexts

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

xual

-0.76

utherford

-0.70

IBLE

-0.66

ioxide

-0.65

 ignorance

-0.64

lled

-0.62

 ministries

-0.61

 charges

-0.61

elaide

-0.60

ouston

-0.60

POSITIVE LOGITS

Kot

1.38

zeb

1.16

aku

0.96

lin

0.91

zon

0.84

ito

0.84

©¶æ¥µ

0.82

estone

0.82

eteenth

0.81

etsu

0.80

Activations Density 0.006%

proper nouns related to a specific name, "Kotick"

references to the name "Kot" in various contexts

No Comments

No Known Activations