INDEX

Explanations

phrases related to time spent or dedicated to certain activities

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

circle

-0.61

 carriers

-0.60

soDeliveryDate

-0.59

 Cricket

-0.58

 leveled

-0.56

 Commando

-0.56

 departures

-0.55

adi

-0.55

rices

-0.53

ells

-0.52

POSITIVE LOGITS

 frivol

0.77

anchester

0.77

 summers

0.77

 waking

0.76

 slee

0.76

borgh

0.74

 wisely

0.73

acebook

0.70

nesota

0.70

baugh

0.69

Activations Density 0.866%

phrases related to time spent or dedicated to certain activities

No Comments

No Known Activations