INDEX

Explanations

literary works and music related terms and titles

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

 referen

-0.97

 carbohyd

-0.81

 skilled

-0.81

 notor

-0.80

aditional

-0.79

 deleg

-0.76

 conflic

-0.75

etheless

-0.74

 unnecess

-0.74

 credential

-0.73

POSITIVE LOGITS

 Productions

1.36

 Syndrome

1.35

âĦ¢

1.25

 Brigade

1.20

 Studios

1.18

 Massacre

1.14

 Reloaded

1.13

 Collective

1.13

!,

1.12

 Records

1.10

Activations Density 1.547%

literary works and music related terms and titles

No Comments

No Known Activations