INDEX

Explanations

sentences related to precaution or warning

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

plex

-0.76

 Explos

-0.60

 sclerosis

-0.60

haul

-0.59

 ingred

-0.57

volume

-0.57

 WORLD

-0.57

Plex

-0.56

lined

-0.56

paces

-0.55

POSITIVE LOGITS

ary

1.66

ARY

1.15

aries

1.14

arily

1.09

autions

0.95

ously

0.94

urous

0.93

arial

0.86

ers

0.85

ari

0.85

Activations Density 0.035%

sentences related to precaution or warning

No Known Activations