INDEX

Explanations

medical terms or descriptions such as injuries or physical health issues

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

terday

-0.74

ivities

-0.70

IVERS

-0.69

imates

-0.69

 ORIG

-0.69

 Balt

-0.69

IVER

-0.66

 nomine

-0.65

IVES

-0.65

IVE

-0.64

POSITIVE LOGITS

bing

1.47

ruary

1.34

bles

1.31

bled

1.25

ulous

1.17

bling

1.15

iotics

1.15

uild

1.15

bed

1.14

urger

1.13

Activations Density 0.935%

medical terms or descriptions such as injuries or physical health issues

No Comments

No Known Activations