INDEX

Explanations

words related to escalating or increasing actions

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

ternity

-0.75

pedia

-0.68

 Samar

-0.68

 circumst

-0.68

creen

-0.67

ĨĴ

-0.66

nsic

-0.66

KNOWN

-0.65

 Confederation

-0.64

ä¸ī

-0.63

POSITIVE LOGITS

aging

1.24

arts

1.07

ages

1.04

up

1.00

antly

0.98

aged

0.98

hetamine

0.95

aign

0.92

inson

0.89

age

0.87

Activations Density 0.019%

words related to escalating or increasing actions

No Comments

No Known Activations