INDEX

Explanations

phrases related to nuclear warfare and international tensions, particularly focused on North Korea and related conflicts

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

tyard

-0.79

haw

-0.75

Row

-0.74

Preview

-0.71

olulu

-0.70

Ort

-0.69

 Agric

-0.69

00200000

-0.69

Tee

-0.69

Marginal

-0.67

POSITIVE LOGITS

 warheads

1.11

 weapons

1.06

 submarines

1.05

 deton

1.05

 missile

1.03

 submarine

1.01

 disarm

1.00

 deterrent

1.00

 reactor

1.00

 bomb

0.98

Activations Density 0.748%

phrases related to nuclear warfare and international tensions, particularly focused on North Korea and related conflicts

No Comments

No Known Activations