INDEX

Explanations

mentions of specific names, likely related to social media accounts or individuals

oai_token-act-pair · gpt-3.5-turbo

proper nouns and names related to individuals or organizations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

zos

-0.90

Rez

-0.89

 Jonas

-0.86

Pok

-0.86

Zen

-0.83

zn

-0.83

zo

-0.82

ten

-0.82

 Eisen

-0.81

Pie

-0.75

POSITIVE LOGITS

iler

0.87

av

0.85

taboola

0.84

ave

0.84

aver

0.83

BB

0.81

aving

0.80

CRIP

0.80

cle

0.78

iling

0.77

Activations Density 0.589%

mentions of specific names, likely related to social media accounts or individuals

proper nouns and names related to individuals or organizations

No Comments

No Known Activations