INDEX

Explanations

text related to role-playing games and character attributes

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

çĶ°

-0.72

sburg

-0.72

ettel

-0.72

Asia

-0.71

Ò

-0.71

Redd

-0.70

ALP

-0.69

markets

-0.67

uminati

-0.67

India

-0.66

POSITIVE LOGITS

Spell

1.20

 enchanted

1.17

 Arcane

1.17

 Spells

1.15

 Spell

1.14

 enchantment

1.09

spell

1.08

nec

1.06

 Haste

1.05

 enchant

1.05

Activations Density 0.808%

text related to role-playing games and character attributes

No Comments

No Known Activations