INDEX

Explanations

terms with the prefix "mon-" followed by a digit indicating intensity or order

oai_token-act-pair · gpt-3.5-turbo

terms related to "monolithic" structures or concepts

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

 Norn

-0.89

éĹĺ

-0.89

ULTS

-0.84

OHN

-0.83

ORTS

-0.79

CRIP

-0.79

æĸ¹

-0.79

ÃĽ

-0.79

IBLE

-0.78

IRO

-0.75

POSITIVE LOGITS

stros

1.24

olithic

1.16

olith

1.13

soon

1.12

oton

1.04

ocular

1.01

omial

0.99

ograph

0.99

ter

0.98

itored

0.97

Activations Density 0.010%

terms with the prefix "mon-" followed by a digit indicating intensity or order

terms related to "monolithic" structures or concepts

No Comments

No Known Activations