INDEX

Explanations

examples of actions or traits that are openly declared or acknowledged

oai_token-act-pair · gpt-3.5-turbo

instances of the word "openly"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

ammy

-0.77

anan

-0.75

 Summoner

-0.73

 Tune

-0.71

oday

-0.70

error

-0.69

wich

-0.69

akeru

-0.67

 Tags

-0.67

oku

-0.66

POSITIVE LOGITS

 openly

1.07

odox

0.83

 trumpet

0.73

 confess

0.72

lique

0.72

entimes

0.71

 disreg

0.71

 profess

0.70

 acknowledged

0.69

 concealed

0.69

Activations Density 0.006%

examples of actions or traits that are openly declared or acknowledged

instances of the word "openly"

No Comments

No Known Activations