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-2-9B
    4. Residual Stream - 16k
    5. 20-GEMMASCOPE-RES-16K
    6. 5
    Prev
    Next
    INDEX
    Explanations

    references to government or authority organizations and their actions

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

    references to years and time periods.

    oai_token-act-pair · claude-3-5-haiku-20241022Triggered by @neilrathi

    Frequent use of prepositions and conjunctions suggests a focus on relationships between ideas and detailed descriptions in various contexts.

    eleuther_acts_top20 · gpt-4o-miniTriggered by @neilrathi

    phrases related to recorded events or notable stories involving specific individuals.

    oai_token-act-pair · gpt-4oTriggered 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
     also
    -0.39
    stdc
    -0.37
     religione
    -0.35
     isso
    -0.34
    esistenza
    -0.34
    ufficio
    -0.32
     <<<<<<<<<<<<<<
    -0.32
     menurut
    -0.31
     itu
    -0.30
    instancetype
    -0.30
    POSITIVE LOGITS
    <unused52>
    0.73
    <unused14>
    0.73
    [@BOS@]
    0.73
    <unused74>
    0.73
    <unused3>
    0.73
    <unused79>
    0.73
    <unused42>
    0.73
    <unused51>
    0.73
    <unused23>
    0.73
    <unused8>
    0.73
    Activations Density 0.504%

    No Known Activations