Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 22-GEMMASCOPE-RES-16K
    6. 3403
    Prev
    Next
    INDEX
    Explanations

    variables and their numerical values in code

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    Code, schemas, and data structures

    np_acts-logits-general · gemini-2.0-flash

    data structure definitions

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 22-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_22/width_16k/average_l0_72
    Prompts (Dashboard)
    36,864 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Hook Layer
    22
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    PubMed
    -0.55
    ἶ
    -0.53
     Meksiku
    -0.53
    lück
    -0.49
     démocratique
    -0.49
     culturelles
    -0.49
     humaines
    -0.47
    koning
    -0.47
    clientHeight
    -0.45
    益
    -0.45
    POSITIVE LOGITS
    SourceChecksum
    1.03
     autorytatywna
    1.03
    ConstraintMaker
    0.98
    MLLoader
    0.84
    SequentialGroup
    0.84
     NSCoder
    0.75
    \{\\
    0.71
    انيف
    0.71
    دانشنامهٔ
    0.70
     CreateTagHelper
    0.70
    Activations Density 0.040%

    No Known Activations