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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    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. 9528
    Prev
    Next
    INDEX
    Explanations

    phrases related to causal connections in legal and workplace contexts

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

    language expressing legal elements and standards in civil/employment tort analysis, especially causation, intent, severity, and liability criteria within judicial opinions and citations.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    legal language describing the establishment or demonstration of claims, evidence, causation, and requirements in formal legal proceedings.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408
    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
     surla
    -0.42
    spo
    -0.41
    conten
    -0.40
     awful
    -0.39
    Vision
    -0.38
     spoilers
    -0.38
    ButterKnife
    -0.37
     instalment
    -0.37
    httphttps
    -0.36
     sundry
    -0.36
    POSITIVE LOGITS
    __":
    
    0.49
    awtextra
    0.47
    crollView
    0.46
     arşivlendi
    0.44
     betweenstory
    0.44
     calendriers
    0.43
    adaptiveStyles
    0.43
    GEBURTSDATUM
    0.43
     للمعارف
    0.40
    Portale
    0.39
    Activations Density 1.551%

    No Known Activations