© 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-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 65002
    Prev
    Next
    INDEX
    Explanations

    obstruction and interference charges

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron fires on words that denote legal or criminal actions and proceedings—e.g. mentions of being charged, convicted, found guilty, indicted, or held in contempt.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     diete
    -0.85
     stings
    -0.80
     expropri
    -0.79
     phosphoric
    -0.76
     Imported
    -0.75
     Harmful
    -0.75
     diphtheria
    -0.75
    𝘐
    -0.72
     bores
    -0.72
     OTO
    -0.71
    POSITIVE LOGITS
     obstruction
    1.84
     tampering
    1.52
     obstruct
    1.43
     tamper
    1.41
     interference
    1.33
     destruction
    1.27
     interfere
    1.24
     destroy
    1.24
     obstructive
    1.23
     obstructing
    1.23
    Activations Density 0.022%

    No Known Activations