INDEX

Explanations

places or countries

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

ext

-0.63

 track

-0.59

 elig

-0.56

 blazing

-0.55

 Scion

-0.54

 fetch

-0.54

 nonexistent

-0.54

-0.53

 consolidation

-0.53

 ISBN

-0.53

POSITIVE LOGITS

 Pradesh

0.88

ilus

0.78

alon

0.78

heim

0.78

ibo

0.75

urnal

0.73

sworth

0.73

ersion

0.73

ilk

0.72

otropic

0.72

Activations Density 1.218%

places or countries

No Comments

No Known Activations