INDEX

Explanations

statements or assertions that are considered to be true or accurate

oai_token-act-pair · gpt-3.5-turbo

assertions of truth or validity

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

uty

-0.76

 Sunder

-0.75

onut

-0.71

ourning

-0.71

Pages

-0.70

uled

-0.69

adish

-0.68

ovie

-0.68

etsk

-0.68

mun

-0.66

POSITIVE LOGITS

 believers

0.79

 believer

0.75

polit

0.73

 è£ıè¦ļéĨĴ

0.73

lla

0.73

éĸ

0.68

 emancipation

0.68

 paternity

0.67

terday

0.67

Ñĭ

0.66

Activations Density 0.016%

statements or assertions that are considered to be true or accurate

assertions of truth or validity

No Comments

No Known Activations