INDEX

Explanations

proper nouns, particularly names and titles

oai_token-act-pair · gpt-3.5-turbo

punctuation and numbers used in various contexts

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

 incent

-0.87

undai

-0.66

 neighb

-0.66

ouble

-0.65

 livest

-0.64

 outsourcing

-0.64

 waive

-0.64

 ende

-0.64

terday

-0.63

 bargain

-0.62

POSITIVE LOGITS

Contents

1.34

rawdownloadcloneembedreportprint

1.15

âĵĺ

1.10

Trivia

1.05

pmwiki

1.03

Synopsis

1.03

Overview

0.95

References

0.94

Usage

0.94

Appearance

0.93

Activations Density 0.233%

proper nouns, particularly names and titles

punctuation and numbers used in various contexts

No Comments

No Known Activations