INDEX

Explanations

phrases related to legal and technical discussions

oai_token-act-pair · gpt-3.5-turbo

punctuation marks, particularly commas, in the text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

²¾

-0.79

Widget

-0.74

ļéĨĴ

-0.74

¶ħ

-0.73

ãĤ©

-0.73

ĻĤ

-0.70

ħĭ

-0.67

ldom

-0.66

´

-0.66

¬¼

-0.63

POSITIVE LOGITS

 wrote

1.41

 reads

1.38

 writes

1.24

 said

1.12

 commented

1.12

 explained

1.09

 explains

1.07

 says

1.06

 stated

1.05

 according

1.04

Activations Density 0.127%

phrases related to legal and technical discussions

punctuation marks, particularly commas, in the text

No Comments

No Known Activations