INDEX

Explanations

culinary-related words, potentially referring to dishes or ingredients

oai_token-act-pair · gpt-3.5-turbo

words with specific character sequences or patterns

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 trainers

-0.67

 writ

-0.67

 mathemat

-0.67

 landmarks

-0.65

etsk

-0.63

 runners

-0.63

eatures

-0.63

 explan

-0.62

 nour

-0.62

 myster

-0.61

POSITIVE LOGITS

ï¸ı

1.22

vernment

0.97

lean

0.92

MQ

0.88

ï¸

0.87

log

0.85

ãĥĥãĥī

0.83

Ģ

0.82

ļ

0.81

leans

0.80

Activations Density 0.035%

culinary-related words, potentially referring to dishes or ingredients

words with specific character sequences or patterns

No Comments

No Known Activations