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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B
    7. Residual Stream - 16k
    9. 31-GEMMASCOPE-RES-16K
    11. 14994
    Prev
    Next
    INDEX
    Explanations

    methods and approaches related to experiments and techniques in scientific research

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

    This neuron identifies words and phrases related to methodologies, techniques, approaches, and protocols used in research or technical processes.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    methodological approaches and technical procedures described in academic or scientific writing.

    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
     vouloir
    -0.34
    OCCURRED
    -0.33
     øns
    -0.33
     препратки
    -0.30
     khoác
    -0.30
     Formación
    -0.30
     placée
    -0.30
    lapsingToolbar
    -0.30
    íritu
    -0.30
     Ausstattung
    -0.30
    POSITIVE LOGITS
    AndEndTag
    0.74
     devised
    0.66
     employed
    0.65
     adopted
    0.65
    adopted
    0.65
     whereby
    0.59
     uſed
    0.58
    employed
    0.57
     méthodes
    0.57
     angewendet
    0.57
    Activations Density 0.771%

    No Known Activations