INDEX

Explanations

words related to cooking and recipes

oai_token-act-pair · gpt-3.5-turbo

references to cookware or cooking utensils

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

åŃ

-0.78

ĳå£«

-0.78

IGH

-0.77

@#&

-0.74

 GEAR

-0.73

 Seym

-0.70

çĭ

-0.67

ãĤī

-0.67

è¦

-0.65

marked

-0.65

POSITIVE LOGITS

ning

1.09

zees

0.95

eways

0.90

xual

0.85

olin

0.83

ned

0.80

psons

0.79

ahon

0.77

oche

0.76

pan

0.76

Activations Density 0.018%

words related to cooking and recipes

references to cookware or cooking utensils

No Comments

No Known Activations