INDEX

Explanations

dates or timeframes

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

ingen

-0.92

lain

-0.89

umen

-0.89

udeb

-0.89

leck

-0.87

ickson

-0.85

ounce

-0.79

ItemTracker

-0.79

ilts

-0.76

YR

-0.74

POSITIVE LOGITS

 relatives

1.00

 distant

0.97

 cousin

0.95

 cousins

0.93

 shores

0.87

 neighbours

0.82

 ancestors

0.81

 kindred

0.80

 neighbour

0.80

 Territories

0.78

Activations Density 0.032%

dates or timeframes

No Comments

No Known Activations