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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    3. Gemma-2-2B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 4993
    Prev
    Next
    INDEX
    Explanations

    scientific and medical terms related to research methods.

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot

    critical evaluations and discussions of medical studies and findings.

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

    phrases indicating the novelty or uniqueness of the authors' knowledge or findings, particularly variations of "to our knowledge" or "for the first time".

    oai_token-act-pair · claude-3-5-sonnet-20240620Triggered by @bot

    "To our/the best of knowledge" phrases

    np_acts-logits-general · gemini-2.0-flash

    best of our knowledge

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GEMMA-2-2B @ 20-gemmascope-res-16k
    Configuration
    google/gemma-scope-2b-pt-res/layer_20/width_16k/average_l0_71
    Prompts (Dashboard)
    36,864 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
    setImageBitmap
    -0.62
    GOTREF
    -0.61
    XmlAccessorType
    -0.60
    evos
    -0.59
     كومونز
    -0.59
    writeFieldEnd
    -0.58
    MigrationBuilder
    -0.56
    Personendaten
    -0.56
    optarg
    -0.55
    toThrow
    -0.54
    POSITIVE LOGITS
     knowledge
    0.90
    knowledge
    0.82
    Rhestr
    0.76
     best
    0.76
     Knowledge
    0.74
     pengetahuan
    0.74
     afa
    0.74
     AFA
    0.74
    AFA
    0.73
     soweit
    0.73
    Activations Density 0.203%

    No Known Activations