INDEX

Explanations

mentions of the name "Obama" and related terms

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_fs12288-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_12288

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

12,288

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

 paperback

-0.71

ishable

-0.66

 charm

-0.66

 flash

-0.64

¥µ

-0.64

izer

-0.63

 whiskey

-0.63

emo

-0.63

ffee

-0.63

 Wicked

-0.62

POSITIVE LOGITS

âĢ

1.82

Â

1.19

 âĢ

1.09

âī

1.09

âĪ

1.09

Â¢

1.08

âĢł

1.07

1.05

Â´

1.04

Activations Density 0.412%

mentions of the name "Obama" and related terms

No Comments

No Known Activations

mentions of the name "Obama" and related terms

No Comments

No Known Activations