INDEX

Explanations

expressions related to updates or announcements within conversations or texts

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

 supposed

-0.79

 supposedly

-0.69

 allegedly

-0.66

urus

-0.63

inus

-0.61

emonic

-0.61

heit

-0.60

Enter

-0.60

uras

-0.59

reated

-0.59

POSITIVE LOGITS

 tomorrow

1.28

 someday

1.21

 hereafter

1.02

 anytime

1.01

 soon

0.95

 sooner

0.87

 whoever

0.85

 sometime

0.85

morrow

0.84

 tonight

0.80

Activations Density 2.412%

expressions related to updates or announcements within conversations or texts

No Comments

No Known Activations