INDEX

Explanations

information related to historical events involving military operations and aircrafts

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

conom

-0.90

geist

-0.85

weet

-0.79

ãĤ¦ãĤ¹

-0.78

 Rowling

-0.77

galitarian

-0.77

 Flav

-0.76

enfranch

-0.74

 Wealth

-0.73

Gamer

-0.71

POSITIVE LOGITS

 reconnaissance

1.71

 helicopters

1.55

 aircraft

1.53

 manned

1.50

 unmanned

1.45

 stationed

1.44

 helicopter

1.40

 airborne

1.40

 submarines

1.38

 convoy

1.36

Activations Density 4.575%

information related to historical events involving military operations and aircrafts

No Comments

No Known Activations