INDEX

Explanations

mentions of the letter "J" preceded by a specific context

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter "J" 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 @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 late

-0.67

 captcha

-0.66

 secondly

-0.63

 peanuts

-0.63

 Anthem

-0.62

 MacArthur

-0.61

 Senegal

-0.61

 DeVos

-0.60

cluding

-0.60

 heads

-0.60

POSITIVE LOGITS

igsaw

1.38

ealous

1.36

argon

1.26

umbo

1.23

ockey

1.23

umper

1.22

UMP

1.20

okers

1.17

ules

1.17

umps

1.16

Activations Density 0.029%

mentions of the letter "J" preceded by a specific context

occurrences of the letter "J" in various contexts

No Comments

No Known Activations