INDEX

Explanations

phrases related to mythology and characters from myths and stories as well as people in legal trouble

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

kefeller

-0.89

acca

-0.86

ype

-0.79

vernment

-0.77

oday

-0.76

aleb

-0.75

resy

-0.73

querque

-0.71

redit

-0.71

don

-0.71

POSITIVE LOGITS

 Anne

1.28

 Marie

1.17

 Louise

1.16

 Elizabeth

1.09

 Mary

1.08

 Isabel

1.08

 Patricia

1.05

 herself

1.05

 Margaret

1.04

 Diana

1.04

Activations Density 0.089%

phrases related to mythology and characters from myths and stories as well as people in legal trouble

No Comments

No Known Activations