INDEX

Explanations

food-related references and mentions of eating

oai_token-act-pair · gpt-3.5-turbo

instances of the word "eat" and variations related to eating

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

Aus

-0.80

 Gemini

-0.75

 Stellar

-0.72

 Lash

-0.71

channelAvailability

-0.71

 Archangel

-0.70

PV

-0.68

 Bowman

-0.68

 McKenna

-0.67

 Nichols

-0.67

POSITIVE LOGITS

 habits

1.11

eteria

1.08

eat

0.91

anut

0.90

 foods

0.89

 eaten

0.89

etics

0.88

esta

0.87

 breakfast

0.87

oleon

0.86

Activations Density 0.023%

food-related references and mentions of eating

instances of the word "eat" and variations related to eating

No Comments

No Known Activations