INDEX

Explanations

references to head coverings or the concept of "head" 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_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

PsyNetMessage

-0.77

Mub

-0.70

 Niet

-0.66

ISM

-0.65

 autos

-0.65

 Chocobo

-0.65

eers

-0.62

CK

-0.62

 Parenthood

-0.60

 recip

-0.60

POSITIVE LOGITS

gear

1.39

dress

1.35

liners

1.31

canon

1.30

quarter

1.29

butt

1.28

liner

1.27

lining

1.26

stock

1.24

band

1.22

Activations Density 0.196%

references to head coverings or the concept of "head" in various contexts

No Comments

No Known Activations