INDEX

Explanations

mentions of food-related terms

oai_token-act-pair · gpt-3.5-turbo

references to food and food-related topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Wink

-0.85

 Caldwell

-0.73

Jed

-0.71

assis

-0.69

owered

-0.69

 Woods

-0.67

Gib

-0.66

âĸ¬

-0.66

 Painter

-0.65

-0.64

POSITIVE LOGITS

stuff

1.17

grain

1.09

borne

1.01

 eaten

0.95

 allergies

0.94

corn

0.91

 additives

0.90

mone

0.90

 intake

0.89

 starch

0.87

Activations Density 0.042%

mentions of food-related terms

references to food and food-related topics

No Comments

No Known Activations