INDEX

Explanations

events or actions taking place in the future

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_fs1536-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

1,536

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

cause

-0.61

mor

-0.58

ums

-0.58

-+

-0.56

him

-0.56

agra

-0.53

into

-0.53

 rightful

-0.53

onto

-0.52

yr

-0.51

POSITIVE LOGITS

 there

1.05

we

0.96

 they

0.93

,.

0.81

 however

0.74

 pandemonium

0.73

it

0.72

 THERE

0.72

he

0.70

Activations Density 2.962%

events or actions taking place in the future

No Comments

No Known Activations