INDEX

Explanations

names of people being targeted or affected by violence or injustice

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

Not in Any Lists

No Comments

Negative Logits

phis

-1.00

-0.82

acular

-0.82

ivas

-0.80

ivery

-0.80

enium

-0.80

ertodd

-0.79

achusetts

-0.79

imates

-0.78

neapolis

-0.78

POSITIVE LOGITS

zz

0.95

zza

0.88

ÄŁ

0.86

ÅŁ

0.85

gger

0.83

Ã±

0.82

Ã§a

0.80

FORE

0.79

zzi

0.76

legates

0.74

Activations Density 0.043%

names of people being targeted or affected by violence or injustice

No Comments

No Known Activations