INDEX

Explanations

dark and ominous descriptions or contexts

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

utic

-0.75

agine

-0.74

ovo

-0.73

raltar

-0.72

Dur

-0.71

cule

-0.71

utable

-0.68

ufact

-0.68

Ĥİ

-0.68

PsyNetMessage

-0.68

POSITIVE LOGITS

ening

1.32

ened

1.12

horse

0.93

eners

0.92

room

0.91

 clouds

0.89

 recess

0.89

haired

0.89

dark

0.85

hound

0.84

Activations Density 0.019%

dark and ominous descriptions or contexts

No Comments

No Known Activations