INDEX

Explanations

details related to human interactions and actions

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

specific nouns and their variations related to characters and objects

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

 FOIA

-0.69

 FILE

-0.60

 Reconstruction

-0.58

 Catalonia

-0.58

 commons

-0.56

 Serie

-0.56

 census

-0.56

 indications

-0.55

BLIC

-0.55

 environment

-0.55

POSITIVE LOGITS

ucker

1.36

uits

1.32

pring

1.29

ucking

1.24

linger

1.21

uit

1.20

ouls

1.19

ayers

1.19

pread

1.18

aber

1.17

Activations Density 0.262%

details related to human interactions and actions

specific nouns and their variations related to characters and objects

No Comments

No Known Activations