INDEX

Explanations

phrases related to violent actions or atrocities

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

BuyableInstoreAndOnline

-0.87

 Benson

-0.65

Scot

-0.63

DCS

-0.62

erson

-0.62

iasm

-0.62

annis

-0.61

 Collider

-0.61

acea

-0.60

agall

-0.60

POSITIVE LOGITS

 slaughtered

1.03

 Massacre

0.99

 massac

0.98

 Slaughter

0.92

houses

0.90

 massacre

0.88

 killings

0.88

quished

0.85

 massacres

0.85

 slaughter

0.84

Activations Density 0.053%

phrases related to violent actions or atrocities

No Comments

No Known Activations