INDEX

Explanations

mentions of the name "Mimi" with varying levels of activation

oai_token-act-pair · gpt-3.5-turbo

the name "mi" as a significant identifier or reference

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

lain

-0.73

rooms

-0.71

swick

-0.71

glers

-0.68

 constants

-0.68

*/(

-0.67

tions

-0.67

 audits

-0.67

tty

-0.64

 Ninth

-0.61

POSITIVE LOGITS

pora

0.97

olla

0.92

ovember

0.87

veland

0.86

mi

0.84

oti

0.83

wi

0.82

imi

0.81

ason

0.79

ota

0.79

Activations Density 0.007%

mentions of the name "Mimi" with varying levels of activation

the name "mi" as a significant identifier or reference

No Comments

No Known Activations