INDEX

Explanations

specific phrases indicating an event or topic is not the first instance

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "the."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

é¾įåĸļå£«

-0.91

inders

-0.77

å§«

-0.74

Maps

-0.72

Background

-0.71

uten

-0.71

Notes

-0.68

pez

-0.68

Brother

-0.67

NOTE

-0.66

POSITIVE LOGITS

 slightest

1.36

 smartest

1.16

 brightest

1.15

 easiest

1.13

 prett

1.07

 same

1.05

 usual

0.99

 safest

0.93

 happiest

0.92

 kind

0.91

Activations Density 0.092%

specific phrases indicating an event or topic is not the first instance

occurrences of the word "the."

No Comments

No Known Activations