INDEX

Explanations

words related to toxic substances or dangerous environmental conditions

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

ĸļ

-0.86

nah

-0.81

different

-0.75

ĪĴ

-0.71

DragonMagazine

-0.70

 ãĤµãĥ¼ãĥĨãĤ£ãĥ¯ãĥ³

-0.68

 Varg

-0.67

 warr

-0.67

eering

-0.65

hole

-0.65

POSITIVE LOGITS

ALLY

1.08

chio

1.00

 acid

1.00

enter

0.98

otine

0.97

ulum

0.94

orp

0.91

ulate

0.90

entric

0.90

hes

0.88

Activations Density 0.034%

words related to toxic substances or dangerous environmental conditions

No Comments

No Known Activations