INDEX

Explanations

words related to social media and online communication

oai_token-act-pair · gpt-3.5-turbo

duplicated characters or symbols

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 semic

-0.81

 scattering

-0.75

 scatter

-0.75

 Dresden

-0.74

 habitable

-0.73

 guiActiveUnfocused

-0.72

 diffusion

-0.68

 confinement

-0.67

 folding

-0.67

 Eisen

-0.66

POSITIVE LOGITS

ª

1.04

¹

1.04

½

0.98

Twitter

0.98

realDonaldTrump

0.96

ı

0.93

°

0.91

ðŁĺ

0.90

¼

0.90

CNN

0.90

Activations Density 0.398%

words related to social media and online communication

duplicated characters or symbols

No Comments

No Known Activations