INDEX

Explanations

detailed descriptions of specific individuals and scenarios related to crime 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_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

 domains

-0.69

uld

-0.69

uers

-0.66

ouls

-0.66

Players

-0.65

plet

-0.65

 increments

-0.64

ãĥ¼ãĥ³

-0.63

 VIDEOS

-0.63

 iPads

-0.63

POSITIVE LOGITS

 extraord

1.33

 specializing

1.22

who

1.18

 whose

1.09

 whom

0.98

who

0.95

ess

0.94

digy

0.91

turned

0.87

 laureate

0.87

Activations Density 1.071%

detailed descriptions of specific individuals and scenarios related to crime events

No Comments

No Known Activations