INDEX

Explanations

German words and names

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

eers

-0.87

states

-0.72

ered

-0.71

quo

-0.70

eer

-0.68

 Hots

-0.66

manship

-0.65

SHIP

-0.64

rers

-0.63

 trumpet

-0.60

POSITIVE LOGITS

udge

1.22

acker

1.18

atton

1.16

anded

1.14

umpy

1.12

acket

1.12

acking

1.10

abbit

1.10

aternity

1.08

agnar

1.07

Activations Density 1.712%

German words and names

No Comments

No Known Activations