INDEX

Explanations

sentences that suggest logical conclusions or connections

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

rongh

-0.73

mage

-0.66

ociate

-0.64

apons

-0.63

-0.62

depend

-0.61

nants

-0.61

iverpool

-0.61

uli

-0.60

è¦

-0.60

POSITIVE LOGITS

 coincidence

0.88

ceivable

0.79

 folly

0.73

 raining

0.69

 hypocritical

0.68

 ironic

0.64

 exaggeration

0.63

EC

0.62

 hypocrisy

0.61

 uphill

0.61

Activations Density 1.769%

sentences that suggest logical conclusions or connections

No Comments

No Known Activations