INDEX

Explanations

sentences starting with "To" followed by verbs that suggest actions or directions

oai_token-act-pair · gpt-3.5-turbo

phrases related to product support and user guidance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 indistinguishable

-0.71

 oath

-0.69

 retweet

-0.63

 grinning

-0.62

 resemblance

-0.61

 quot

-0.60

 TRUMP

-0.59

 smiling

-0.59

illance

-0.58

 Seal

-0.58

POSITIVE LOGITS

Luckily

1.56

Solution

1.52

Fortunately

1.48

Thankfully

1.34

 Luckily

1.32

 Fortunately

1.19

 Solution

1.18

Nevertheless

1.09

Therefore

1.08

 Thankfully

1.07

Activations Density 0.981%

sentences starting with "To" followed by verbs that suggest actions or directions

phrases related to product support and user guidance

No Comments

No Known Activations