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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 4436
    Prev
    Next
    INDEX
    Explanations

    repeated patterns and structured data representations

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

    technical or computational terminology and symbols.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-pt-res/layer_20/width_16k/average_l0_68
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.20.hook_resid_post
    Hook Layer
    20
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     hadas
    -0.37
     Decke
    -0.35
     przecież
    -0.34
     Begegn
    -0.34
    >{@
    -0.34
     azonban
    -0.32
    uttavia
    -0.32
     esferas
    -0.31
     createSlice
    -0.31
     chiens
    -0.31
    POSITIVE LOGITS
    AnchorStyles
    0.69
     disambiguazione
    0.59
    AxisAlignment
    0.56
    CloseOperation
    0.55
     CreateTagHelper
    0.54
     noDo
    0.54
    帖最后由
    0.52
    الحياه
    0.52
     onBind
    0.49
    niająca
    0.48
    Activations Density 0.581%

    No Known Activations