INDEX

Explanations

adjectives or descriptive terms for various qualities

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_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

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

Downloadha

-0.83

 elig

-0.81

 dictated

-0.79

itism

-0.79

 socially

-0.79

 estab

-0.78

 administr

-0.77

 harshly

-0.76

 diluted

-0.76

 duly

-0.75

POSITIVE LOGITS

 Skies

1.30

 Ones

1.24

 Bastard

1.19

 Monkey

1.17

 Intent

1.15

 Birds

1.13

 Creatures

1.13

 Beasts

1.10

 Realms

1.09

 Thoughts

1.07

Activations Density 0.156%

adjectives or descriptive terms for various qualities

No Comments

No Known Activations