INDEX

Explanations

references to wisdom and wise concepts 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

akedown

-0.73

ching

-0.72

leased

-0.72

ches

-0.68

rary

-0.63

dule

-0.63

ched

-0.61

properties

-0.61

ãĥĺãĥ©

-0.61

alien

-0.60

POSITIVE LOGITS

 wisdom

0.92

fulness

0.88

smanship

0.87

 glean

0.78

 Advice

0.77

 curve

0.74

isdom

0.74

 quot

0.70

uracy

0.69

 regarding

0.67

Activations Density 0.007%

references to wisdom and wise concepts in various contexts

No Comments

No Known Activations