INDEX

Explanations

mentions of political activities, decision-making, and policies

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 lodged

-0.72

NOW

-0.69

%%

-0.67

llor

-0.64

killed

-0.63

0000000000000000

-0.62

CLASSIFIED

-0.62

username

-0.62

HAEL

-0.61

casters

-0.60

POSITIVE LOGITS

 lieu

1.59

 accordance

1.49

 conjunction

1.45

effic

1.33

 spite

1.33

 relation

1.29

clusions

1.28

efficiency

1.28

 vitro

1.27

ordinate

1.27

Activations Density 2.240%

mentions of political activities, decision-making, and policies

No Comments

No Known Activations