INDEX

Explanations

mentions of locations or geopolitical topics, specifically around Russia and Crimea

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

IFrame

Link

Negative Logits

oples

-0.72

 heartbeat

-0.66

heric

-0.65

wcs

-0.64

hetical

-0.62

 rall

-0.61

ļéĨĴ

-0.61

 restrooms

-0.60

STER

-0.60

 Reloaded

-0.59

POSITIVE LOGITS

nesses

1.01

liest

0.96

ness

0.94

ings

0.91

onest

0.90

washer

0.82

imaru

0.82

»Ĵ

0.82

ments

0.80

lihood

0.80

Activations Density 0.029%

mentions of locations or geopolitical topics, specifically around Russia and Crimea

No Known Activations