INDEX

Explanations

action-related words and phrases related to planning and organization

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

shire

-0.65

inas

-0.65

Jr

-0.65

avery

-0.62

ailable

-0.61

Cause

-0.60

 charism

-0.60

 Trophy

-0.60

 Jude

-0.60

 Stain

-0.59

POSITIVE LOGITS

 Parenthood

0.98

etary

0.90

paren

0.86

isphere

0.85

ters

0.82

escape

0.82

emaker

0.81

obs

0.78

 plans

0.77

 accordingly

0.75

Activations Density 1.657%

action-related words and phrases related to planning and organization

No Comments

No Known Activations