INDEX

Explanations

phrases indicating a list or continuation of information

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

mand

-0.63

atorium

-0.60

glas

-0.57

 intrusion

-0.57

 informing

-0.57

 deposit

-0.57

autions

-0.55

 informed

-0.55

 respect

-0.55

itled

-0.55

POSITIVE LOGITS

etc

1.62

etc

1.20

 Lastly

1.11

 Finally

1.00

 respectively

0.97

anything

0.95

ect

0.91

whatever

0.88

These

0.86

Lastly

0.86

Activations Density 2.218%

phrases indicating a list or continuation of information

No Comments

No Known Activations