INDEX

Explanations

mentions of the word "books"

oai_token-act-pair · gpt-3.5-turbo

references to books and literature

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

mod

-0.68

bug

-0.68

 prog

-0.67

 sens

-0.64

 convention

-0.64

 Tanner

-0.64

 lodged

-0.64

 vent

-0.64

 neon

-0.63

 prec

-0.62

POSITIVE LOGITS

books

3.76

Books

1.88

times

1.50

ebook

1.34

book

1.28

papers

1.28

atches

1.14

cards

1.14

cases

1.12

 Journals

1.11

Activations Density 0.028%

mentions of the word "books"

references to books and literature

No Comments

No Known Activations