INDEX

Explanations

descriptions related to doors and entryways

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_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

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

irtual

-0.76

utan

-0.75

wana

-0.69

elson

-0.67

 Rounds

-0.67

 Capital

-0.67

HER

-0.67

inational

-0.67

 Asset

-0.66

ocious

-0.65

POSITIVE LOGITS

 doors

1.41

door

1.28

 door

1.23

 closet

1.04

 Door

1.01

 doorway

1.01

 upstairs

1.01

 downstairs

1.01

doors

0.99

 renovation

0.98

Activations Density 1.259%

descriptions related to doors and entryways

No Comments

No Known Activations