© 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-IT
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 13783
    Prev
    Next
    INDEX
    Explanations

    specific data fields and values in structured formats, indicating serialization and deserialization operations

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

    Values, often with colons and storage references

    np_acts-logits-general · gemini-2.0-flash

    code data structures

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-9B-IT @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_16k/average_l0_91
    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
     queſta
    -0.85
    BufferException
    -0.82
    webElementXpaths
    -0.80
    AndroidJUnit
    -0.79
     laſſen
    -0.77
    .}~\
    -0.77
    <unused23>
    -0.76
    <unused52>
    -0.76
    <unused14>
    -0.76
    <pad>
    -0.76
    POSITIVE LOGITS
    =
    0.30
    :
    0.29
    styleUrls
    0.27
    e
    0.25
    łość
    0.25
    czki
    0.23
     =
    0.23
     utama
    0.23
    wną
    0.23
    figsize
    0.23
    Activations Density 0.445%

    No Known Activations