INDEX

Explanations

words related to data security breaches and financial information

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

 Tycoon

-0.96

 Oaks

-0.79

 Grill

-0.71

 Grind

-0.70

Dug

-0.70

 Showdown

-0.69

 Spoon

-0.68

 Banner

-0.66

 Franks

-0.65

 Gardens

-0.65

POSITIVE LOGITS

linear

1.20

chal

1.18

verbal

1.14

fiction

1.12

profit

1.11

zero

1.10

stop

1.10

technical

1.08

threatening

1.08

existent

1.08

Activations Density 0.280%

words related to data security breaches and financial information

No Comments

No Known Activations