INDEX

Explanations

numerical values indicating different options or versions available for purchase

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

awaru

-0.73

 Lumpur

-0.65

orative

-0.64

 mascul

-0.60

 propri

-0.60

 hygiene

-0.58

 Pompe

-0.58

 sway

-0.58

enges

-0.55

ategory

-0.55

POSITIVE LOGITS

st

1.44

Password

1.34

1.27

1.26

1.13

1.09

Â½

1.03

âģ

1.01

0.99

0.94

Activations Density 0.407%

numerical values indicating different options or versions available for purchase

No Comments

No Known Activations