© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 57343
    Prev
    Next
    INDEX
    Explanations

    ly

    np_max-act · gemini-2.0-flash

    This neuron detects formal technical specification phrases common in patent text—words like “for use in,” “comprising,” “adaptable to,” and similar claim‐style verbs and modifiers.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    apat
    -0.07
    .Av
    -0.06
     są
    -0.06
    _DIS
    -0.06
    ％
    -0.06
    ！！
    -0.06
    오
    -0.06
     :/
    -0.06
    .AR
    -0.06
    arseille
    -0.06
    POSITIVE LOGITS
     retire
    0.07
    cae
    0.07
    .Geometry
    0.07
    	INNER
    0.07
    _rsp
    0.07
    or
    0.06
     vers
    0.06
    競
    0.06
    detail
    0.06
     angled
    0.06
    Activations Density 0.090%

    No Known Activations