INDEX

Explanations

phrases related to loss or danger to human life

oai_token-act-pair · gpt-3.5-turbo

references to human lives and their significance in various contexts, particularly those involving risk or loss

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

CAST

-0.70

ane

-0.70

NetMessage

-0.64

atorial

-0.62

Marginal

-0.62

ripp

-0.60

ractive

-0.60

ggles

-0.59

IDE

-0.58

iasis

-0.58

POSITIVE LOGITS

 lives

0.86

chool

0.84

journal

0.83

lihood

0.82

sole

0.81

©¶æ

0.77

 Forever

0.76

ynthesis

0.74

pun

0.73

bage

0.72

Activations Density 0.014%

phrases related to loss or danger to human life

references to human lives and their significance in various contexts, particularly those involving risk or loss

No Comments

No Known Activations