INDEX

Explanations

phrases related to negative attributes or concepts

oai_token-act-pair · gpt-3.5-turbo

negative phrases or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

minster

-0.73

realDonaldTrump

-0.65

lycer

-0.64

iership

-0.62

ual

-0.57

ulative

-0.56

ousand

-0.55

RAFT

-0.55

right

-0.55

 simplicity

-0.53

POSITIVE LOGITS

xious

1.33

ct

1.06

obs

1.05

va

1.04

vae

0.98

vel

0.96

onday

0.96

 longer

0.95

otrop

0.94

zzle

0.93

Activations Density 0.075%

phrases related to negative attributes or concepts

negative phrases or conditions

No Comments

No Known Activations