INDEX

Explanations

vocabulary related to political news and events

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ABE

-0.70

 Tasman

-0.70

 Niet

-0.70

 Shelter

-0.67

RAD

-0.65

ilers

-0.65

 Marian

-0.64

Wad

-0.64

 Ascension

-0.64

 Doodle

-0.64

POSITIVE LOGITS

Ĳ

1.15

ª

1.13

appropriately

1.05

almost

1.05

ĳ

1.03

¹

1.02

«

1.02

¤

1.01

Ķ

1.00

ı

1.00

Activations Density 0.138%

vocabulary related to political news and events

No Comments

No Known Activations