INDEX

Explanations

mentions of performing or participating in 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_fs768-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_768

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

768

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

 background

-0.66

 marsh

-0.65

dex

-0.65

 scorp

-0.65

 swamp

-0.65

 foreground

-0.64

 colon

-0.64

arm

-0.64

 evening

-0.63

stem

-0.63

POSITIVE LOGITS

 Nope

1.42

 Answer

1.36

 Well

1.33

 Surely

1.30

Answer

1.29

 Probably

1.24

 Certainly

1.24

 Wouldn

1.22

 Would

1.20

Why

1.18

Activations Density 2.095%

mentions of performing or participating in activities

No Comments

No Known Activations