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-IT
    4. Residual Stream - 131k
    5. 20-GEMMASCOPE-RES-131K
    6. 35424
    Prev
    Next
    INDEX
    Explanations

    architectural and structural features of buildings and objects

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

    parts of structures and creatures

    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-131k
    Configuration
    google/gemma-scope-9b-it-res/layer_20/width_131k/average_l0_81
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    131,072
    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
    tvguidetime
    -0.67
    Vidite
    -0.56
     ujednoznacz
    -0.52
     manta
    -0.52
    çade
    -0.51
    IUrlHelper
    -0.49
     Biôgrafia
    -0.49
     Manta
    -0.48
     brightest
    -0.48
     Reſ
    -0.47
    POSITIVE LOGITS
    hyrchwyd
    0.46
    enterOuterAlt
    0.42
     escalera
    0.41
    Portály
    0.39
     sytu
    0.37
    PreInfinity
    0.37
    Hozzáférés
    0.36
    UseVisualStyle
    0.36
     hospod
    0.35
     leiding
    0.35
    Activations Density 0.010%

    No Known Activations