© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 31-GEMMASCOPE-RES-16K
    6. 4793
    Prev
    Next
    INDEX
    Explanations

    technical terms and indicators related to computational processes and results

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

    words and phrases related to computational performance and efficiency metrics.

    oai_token-act-pair · claude-4-5-haikuTriggered by @emiglarou
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 31-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_31/width_16k/average_l0_114
    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
    31
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    istungen
    -0.53
    ########.
    -0.46
    adaptiveStyles
    -0.40
    Referencies
    -0.39
     Anspruch
    -0.36
    gungen
    -0.36
    iconductors
    -0.35
    SequentialGroup
    -0.35
     BrowserModule
    -0.35
    Enllaces
    -0.35
    POSITIVE LOGITS
     filmmaking
    0.63
     viewing
    0.62
     operational
    0.61
     imaging
    0.60
     manufacturing
    0.60
     playback
    0.60
     purchasing
    0.59
     developmental
    0.59
     cooking
    0.58
     computational
    0.58
    Activations Density 1.675%

    No Known Activations