INDEX

Explanations

numeric values and version numbers in code references

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

Version numbers (e.g., 4.0)

np_acts-logits-general · gemini-2.0-flash

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GEMMA-2-9B-IT @ 31-gemmascope-res-16k

Configuration

google/gemma-scope-9b-it-res/layer_31/width_16k/average_l0_76

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Features

16,384

Data Type

float32

Hook Name

blocks.31.hook_resid_post

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

 Steele

-0.42

 Shat

-0.40

idey

-0.40

Rea

-0.40

Rea

-0.39

Nes

-0.39

Ent

-0.38

 архивлан

-0.38

 ange

-0.38

 Stony

-0.37

POSITIVE LOGITS

 AssemblyCulture

0.50

 ModelExpression

0.48

 Corners

0.47

 corners

0.47

afficheront

0.47

 kasarigan

0.45

ospels

0.45

corners

0.44

isContained

0.44

fifths

0.43

Activations Density 0.517%

numeric values and version numbers in code references

Version numbers (e.g., 4.0)

No Comments

No Known Activations