INDEX

Explanations

titles of creative works and their thematic elements

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

 referen

-1.18

 notor

-0.83

 skilled

-0.82

 satell

-0.78

 knowledgeable

-0.77

 carbohyd

-0.77

 millenn

-0.76

matically

-0.75

 credential

-0.74

PDATE

-0.72

POSITIVE LOGITS

 Syndrome

1.37

 Productions

1.11

 Tatt

1.06

 Records

1.01

 Orchestra

1.00

 Studios

0.98

 Massacre

0.98

 Disorder

0.97

 Theory

0.96

 Designs

0.95

Activations Density 0.134%

titles of creative works and their thematic elements

No Comments

No Known Activations