INDEX

Explanations

phrases related to authority, legal systems, and societal issues

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

 Mant

-0.65

 Federation

-0.61

Moz

-0.59

 Seeking

-0.59

 Rising

-0.57

 Lauder

-0.57

 rejection

-0.57

 Brent

-0.57

ele

-0.56

BAL

-0.56

POSITIVE LOGITS

't

1.87

berra

1.29

adian

1.28

NOT

1.25

 afford

1.17

isters

1.11

 easily

1.08

ister

1.01

 safely

0.99

 attest

0.97

Activations Density 0.853%

phrases related to authority, legal systems, and societal issues

No Comments

No Known Activations