INDEX

Explanations

phrases related to shared concepts or responsibilities

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

paio

-0.78

ī

-0.75

annis

-0.73

ichick

-0.71

icans

-0.70

trak

-0.68

onies

-0.67

¶ħ

-0.67

 desper

-0.67

igers

-0.66

POSITIVE LOGITS

Instance

0.78

cro

0.77

itarian

0.76

holders

0.75

 custody

0.73

owners

0.72

 spaces

0.71

holder

0.71

hare

0.71

ership

0.71

Activations Density 0.032%

phrases related to shared concepts or responsibilities

No Comments

No Known Activations