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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-2B
    5. 22-RES-MATRYOSHKA-DC
    7. 452
    Prev
    Next
    INDEX
    Explanations

    words that precede or follow the word "would", and other language regarding potential impacts, benefits, risks, or judgements.

    oai_token-act-pair · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    chanind/gemma-2-2b-batch-topk-matryoshka-saes-w-32k-l0-40/standard/blocks.22.hook_resid_post
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    float32
    Hook Name
    blocks.22.hook_resid_post
    Architecture
    jumprelu
    Context Size
    1,024
    Dataset
    chanind/pile-uncopyrighted-gemma-1024-abbrv-1B
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    脚注の使い方
    -0.77
    BindView
    -0.65
     Worse
    -0.59
     урна
    -0.57
    ьаж
    -0.56
     esternos
    -0.55
    häng
    -0.54
     worse
    -0.54
     urme
    -0.54
    ArrowToggle
    -0.53
    POSITIVE LOGITS
     helps
    1.76
     help
    1.58
     enable
    1.55
     enables
    1.53
    helps
    1.50
     allows
    1.49
     Helps
    1.46
     allow
    1.45
     helping
    1.40
    Helps
    1.38
    Activations Density 4.128%

    No Known Activations