INDEX

Explanations

statements and announcements related to official events or news

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

skilled

-0.74

 underestimated

-0.70

entimes

-0.69

 sucker

-0.68

 experimented

-0.68

liest

-0.67

odynam

-0.67

 Penet

-0.65

 pract

-0.63

 unsus

-0.62

POSITIVE LOGITS

 spokesperson

0.78

 statement

0.78

 apologizing

0.77

 thanking

0.77

Statement

0.75

:"

0.75

irming

0.75

 apology

0.73

 spokeswoman

0.73

 thanked

0.69

Activations Density 0.212%

statements and announcements related to official events or news

No Comments

No Known Activations