INDEX

Explanations

terms and phrases related to informal language forms and euphemisms

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

dinand

-0.77

olphin

-0.75

gio

-0.71

hardt

-0.69

itor

-0.69

Tickets

-0.67

 Democr

-0.66

negie

-0.66

galitarian

-0.65

 Enlight

-0.64

POSITIVE LOGITS

 shorthand

1.12

 terminology

0.96

 phrases

0.92

 jargon

0.92

 phrase

0.89

 coined

0.88

isms

0.84

 language

0.83

 diction

0.82

 term

0.82

Activations Density 0.015%

terms and phrases related to informal language forms and euphemisms

No Comments

No Known Activations