INDEX

Explanations

references to mathematical symbols or expressions, particularly those involving matrices

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-2B @ 0-gemmascope-mlp-65k

Configuration

google/gemma-scope-2b-pt-mlp/layer_0/width_65k/average_l0_72

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

65,536

Data Type

float32

Hook Name

blocks.0.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

ij

-1.49

AndEndTag

-0.65

MemoryWarning

-0.64

IJ

-0.62

modelBuilder

-0.62

canestro

-0.62

ik

-0.62

//

-0.61

ijan

-0.61

POSITIVE LOGITS

 itſelf

1.28

 myſelf

1.02

 Jefus

0.98

 Cæsar

0.95

 Monfieur

0.92

 pleaſure

0.90

ſelves

0.84

 themſelves

0.83

 theſe

0.83

Efq

0.82

Activations Density 0.018%

references to mathematical symbols or expressions, particularly those involving matrices

No Comments

No Known Activations