INDEX

Explanations

terms related to games and card games

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

 GOODMAN

-0.85

 Noon

-0.75

Ll

-0.73

 Line

-0.71

Hon

-0.69

 Ahead

-0.69

 Lack

-0.69

Wee

-0.68

 Grizz

-0.68

 DRAG

-0.66

POSITIVE LOGITS

ological

1.28

ologically

1.24

ical

1.22

onic

1.17

ically

1.16

omorph

1.16

ology

1.14

ologist

1.10

ologists

1.09

ophobic

1.08

Activations Density 0.107%

terms related to games and card games

No Comments

No Known Activations