INDEX

Explanations

phrases related to asking questions or requests

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

interactions and inquiries between individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

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

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_post

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

edition

-0.61

VIDIA

-0.59

Associated

-0.59

ushima

-0.57

ãĥ¼ãĥĨ

-0.55

 earthqu

-0.55

vantage

-0.55

absor

-0.54

Introduced

-0.54

jam

-0.53

POSITIVE LOGITS

why

1.28

 whether

1.26

if

1.15

 questions

1.14

 what

1.14

how

1.13

 about

0.94

WHY

0.94

 permission

0.93

why

0.91

Activations Density 0.117%

phrases related to asking questions or requests

interactions and inquiries between individuals

No Comments

No Known Activations