INDEX

Explanations

references to psychic phenomena

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

ities

-0.75

esson

-0.74

icio

-0.73

animous

-0.69

ciating

-0.69

icably

-0.67

iture

-0.67

ity

-0.66

ifer

-0.64

itarian

-0.63

POSITIVE LOGITS

STEM

1.11

PE

0.99

CLE

0.97

FACE

0.84

ORK

0.83

OND

0.78

MN

0.77

ramid

0.77

ELL

0.76

ank

0.76

Activations Density 0.032%

references to psychic phenomena

No Comments

No Known Activations