INDEX

Explanations

phrases related to personal decisions or choices

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

ATURES

-0.85

iffs

-0.75

casters

-0.74

marks

-0.72

stars

-0.71

reports

-0.71

ometers

-0.69

killers

-0.69

Americans

-0.68

avers

-0.68

POSITIVE LOGITS

 nutshell

1.40

 hurry

1.19

 manner

1.11

 vein

1.02

 crowded

0.99

 classroom

0.94

 timely

0.93

 vain

0.92

 heartbeat

0.92

 vacuum

0.90

Activations Density 0.154%

phrases related to personal decisions or choices

No Comments

No Known Activations