INDEX

Explanations

list items or reasons with numbers indicating importance or order

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

lator

-0.86

ymph

-0.82

iece

-0.72

wana

-0.71

aughed

-0.68

 Roller

-0.64

wrapper

-0.64

robe

-0.62

room

-0.61

 exchange

-0.60

POSITIVE LOGITS

soever

0.95

 exactly

0.77

why

0.76

abl

0.74

 they

0.70

why

0.70

 people

0.67

abouts

0.66

WHY

0.64

we

0.63

Activations Density 0.032%

list items or reasons with numbers indicating importance or order

No Comments

No Known Activations