INDEX

Explanations

specific phrases and sentences related to legal cases, political discussions, and personal beliefs

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

Topics

-0.77

NetMessage

-0.71

Topic

-0.70

 âĵĺ

-0.67

ãĥ³

-0.62

NX

-0.61

ãĤ§

-0.61

favorite

-0.60

folios

-0.59

Page

-0.58

POSITIVE LOGITS

 somehow

0.91

 disguise

0.73

 sufficiently

0.73

 willfully

0.72

 explicitly

0.72

 expressly

0.71

apeshifter

0.71

 oneself

0.70

 physically

0.70

 excuse

0.69

Activations Density 0.237%

specific phrases and sentences related to legal cases, political discussions, and personal beliefs

No Comments

No Known Activations