INDEX

Explanations

words related to physical injury or harm

oai_token-act-pair · gpt-3.5-turbo

references to death or mortality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

gau

-0.69

ger

-0.68

conn

-0.66

cloth

-0.65

ative

-0.63

ging

-0.63

HUD

-0.62

²¾

-0.62

Lens

-0.59

Ze

-0.59

POSITIVE LOGITS

ATH

1.10

laus

0.96

YN

0.96

anwhile

0.91

ONDON

0.89

pmwiki

0.86

racuse

0.85

acus

0.82

EVA

0.82

ERE

0.76

Activations Density 0.022%

words related to physical injury or harm

references to death or mortality

No Comments

No Known Activations