© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Google DeepMind · Exploring Gemma 2 with Gemma Scope
    5. Gemma-2-9B-IT
    7. Residual Stream - 131k
    9. 20-GEMMASCOPE-RES-131K
    11. 110361
    Prev
    Next
    INDEX
    Explanations

    phrases related to legal pleas, especially guilty and not guilty

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

    pleading guilty or not guilty

    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
    +#+#
    -0.64
     estekak
    -0.55
     otomatig
    -0.53
    RTLR
    -0.52
    ]")]
    -0.51
    Kanpo
    -0.49
    NameInMap
    -0.47
     OnTriggerEnter
    -0.46
    typeDefinition
    -0.45
    Autoritní
    -0.45
    POSITIVE LOGITS
    RTCK
    0.39
    Życiorys
    0.38
     nakalista
    0.37
     neither
    0.36
     Defense
    0.35
     Mixed
    0.35
    Neither
    0.34
    neither
    0.34
     defense
    0.33
    balleur
    0.33
    Activations Density 0.034%

    No Known Activations