INDEX

Explanations

titles or instructions starting with "How"

oai_token-act-pair · gpt-3.5-turbo

phrases beginning with "How" that imply instructions or guidance

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

usher

-0.70

 Tanz

-0.68

uthor

-0.66

ivity

-0.65

IFIED

-0.65

 elimination

-0.63

 hereafter

-0.61

peak

-0.59

mor

-0.57

 pont

-0.56

POSITIVE LOGITS

soever

0.95

lers

0.86

itzer

0.82

ling

0.77

ells

0.74

ever

0.73

links

0.72

umbai

0.70

abouts

0.70

 Steps

0.69

Activations Density 0.055%

titles or instructions starting with "How"

phrases beginning with "How" that imply instructions or guidance

No Comments

No Known Activations