INDEX

Explanations

political and social issues related to justice, government policies, and public figures

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

 favor

-0.79

 reflex

-0.78

 favour

-0.74

 morph

-0.74

 territorial

-0.73

 synerg

-0.71

 division

-0.70

 grazing

-0.68

 superior

-0.67

 vacation

-0.67

POSITIVE LOGITS

ï¸ı

1.18

Because

1.02

sure

1.00

there

0.99

Secondly

0.98

ccording

0.95

Therefore

0.94

It

0.94

FORE

0.93

Whereas

0.93

Activations Density 0.971%

political and social issues related to justice, government policies, and public figures

No Comments

No Known Activations