INDEX

Explanations

words related to relationships and connections

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

 Dragonbound

-0.68

BACK

-0.66

LIA

-0.63

BALL

-0.62

abouts

-0.62

waves

-0.61

arella

-0.61

iasis

-0.61

 Carnage

-0.60

bags

-0.59

POSITIVE LOGITS

ernel

0.87

ables

0.83

erving

0.76

ailability

0.74

QL

0.73

intern

0.73

uthor

0.72

pec

0.72

ould

0.70

ãĥŃ

0.70

Activations Density 0.021%

words related to relationships and connections

No Comments

No Known Activations