INDEX

Explanations

references to programming concepts and updates in code functionality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 22-gemmascope-mlp-16k

Configuration

google/gemma-scope-2b-pt-mlp/layer_22/width_16k/average_l0_121

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.22.hook_mlp_out

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

monology/pile-uncopyrighted

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 שוליים

-0.78

Occurrences

-0.65

 kasarigan

-0.57

domésticos

-0.55

juli

-0.55

omeness

-0.54

navigate

-0.53

————————————

-0.53

 Shakspeare

-0.52

esistenza

-0.52

POSITIVE LOGITS

WireFormatLite

0.57

 explicitly

0.51

 recursively

0.48

WireFormat

0.46

itoria

0.46

joba

0.45

].(

0.44

 trou

0.44

 unters

0.42

 reliably

0.42

Activations Density 0.605%

references to programming concepts and updates in code functionality

No Comments

No Known Activations