INDEX

Explanations

mentions of coffee-related terms

oai_token-act-pair · gpt-3.5-turbo

mentions of coffee and related establishments

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

ROR

-0.82

orse

-0.79

oral

-0.72

alez

-0.69

ItemTracker

-0.68

guided

-0.68

ewski

-0.67

umbered

-0.67

igators

-0.67

ORS

-0.66

POSITIVE LOGITS

 beans

1.22

 coffee

1.15

 Beans

1.13

 bean

1.07

 Coffee

1.03

 brewed

0.96

 cafÃ©

0.95

 cocoa

0.95

 cafe

0.95

 coff

0.94

Activations Density 0.008%

mentions of coffee-related terms

mentions of coffee and related establishments

No Comments

No Known Activations