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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 28-GEMMASCOPE-RES-16K
    6. 4082
    Prev
    Next
    INDEX
    Explanations

    comparative phrases emphasizing the degree of something, particularly focused on "more" and "less."

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

    Preceding "more" and describing a quality

    np_acts-logits-general · gemini-2.0-flash

    more followed by adjective

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 28-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_28/width_16k/average_l0_119
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.28.hook_resid_post
    Hook Layer
    28
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     harika
    -0.38
     wundersch
    -0.38
     prachtig
    -0.36
     maravillosa
    -0.35
     prachtige
    -0.35
    ấc
    -0.33
     énorme
    -0.33
    <eos>
    -0.33
     muchísimo
    -0.33
     maravilloso
    -0.33
    POSITIVE LOGITS
    
    0.72
     betweenstory
    0.66
     mergeFrom
    0.65
    ⟬
    0.64
     ProtoMessage
    0.63
    canestro
    0.62
    الحياه
    0.61
    ſelves
    0.60
    RTGC
    0.60
    Tikang
    0.60
    Activations Density 0.187%

    No Known Activations