INDEX

Explanations

phrases related to terrorist activities and individuals

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

galitarian

-0.79

dit

-0.75

 Dolphin

-0.73

âĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢâĶĢ

-0.70

ðĿ

-0.69

 Hearth

-0.69

 Constantin

-0.67

FACE

-0.66

aird

-0.64

 Wind

-0.63

POSITIVE LOGITS

abad

1.14

 terrorists

1.08

 bombings

1.06

 terrorist

1.05

 bombing

1.02

 terrorism

0.97

 bombers

0.97

 jihad

0.94

ISIS

0.93

 mastermind

0.92

Activations Density 0.779%

phrases related to terrorist activities and individuals

No Comments

No Known Activations