INDEX

Explanations

keywords related to political events and investigations

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

 hears

-0.70

ALLY

-0.70

Õ

-0.67

less

-0.66

Args

-0.64

ded

-0.62

INAL

-0.62

ISH

-0.62

\\\\\\\\\\\\\\\\

-0.62

lich

-0.59

POSITIVE LOGITS

pring

1.59

heet

1.48

cape

1.47

pace

1.44

mith

1.41

paces

1.38

cale

1.34

creen

1.31

hift

1.31

peed

1.26

Activations Density 2.141%

keywords related to political events and investigations

No Comments

No Known Activations