INDEX

Explanations

references to thumbs in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

afort

-0.77

 Occupations

-0.72

assetsadobe

-0.72

 Reborn

-0.72

places

-0.72

mingham

-0.67

Wars

-0.65

 Byrne

-0.65

iveness

-0.64

 Kuwait

-0.64

POSITIVE LOGITS

 thumb

1.06

prints

0.99

hole

0.99

print

0.99

stick

0.95

sticks

0.92

eenth

0.90

pad

0.90

tip

0.86

 thumbs

0.83

Activations Density 0.004%

references to thumbs in various contexts

No Comments

No Known Activations