INDEX

Explanations

phrases indicating quantities or measurements

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

hett

-0.67

halla

-0.66

Â¶

-0.65

advertisement

-0.55

sth

-0.55

agher

-0.54

Pg

-0.54

stab

-0.53

 ende

-0.53

Downloadha

-0.53

POSITIVE LOGITS

 sorts

1.42

 varying

1.38

course

1.15

 course

1.14

 dubious

1.03

 differing

1.02

 questionable

1.02

 interest

0.97

 theirs

0.97

 comparable

0.96

Activations Density 0.581%

phrases indicating quantities or measurements

No Comments

No Known Activations