INDEX

Explanations

interrogative statements starting with "What" or "What the hell" asking about actions or behaviors

oai_token-act-pair · gpt-3.5-turbo

questions and exclamations related to actions or activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

 wisely

-0.73

 territ

-0.70

 responsibly

-0.69

unal

-0.68

 convinc

-0.68

 swiftly

-0.65

 markedly

-0.64

 undeniably

-0.64

 aggressively

-0.63

 regularly

-0.63

POSITIVE LOGITS

 gonna

0.94

?!"

0.90

?!

0.88

!?

0.82

!?"

0.82

 fuss

0.80

âĹ¼

0.76

 supposed

0.75

 laughing

0.72

...?

0.71

Activations Density 0.104%

interrogative statements starting with "What" or "What the hell" asking about actions or behaviors

questions and exclamations related to actions or activities

No Comments

No Known Activations