© 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
    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. 56291
    Prev
    Next
    INDEX
    Explanations

    appropriate

    np_max-act · gemini-2.0-flash

    The neuron activates on the “is … appropriate … for” phrase in the non-publication/precedent boilerplate (e.g. “This disposition is not appropriate for publication and is not precedent”).

    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
     podem
    -0.06
    ndef
    -0.06
    =view
    -0.06
    ωμά
    -0.06
    маз
    -0.06
     impartial
    -0.06
     Neil
    -0.06
     Lynn
    -0.06
    	rec
    -0.06
    (visible
    -0.06
    POSITIVE LOGITS
     Null
    0.08
     gest
    0.07
    east
    0.07
    _Response
    0.07
     sera
    0.07
    "${
    0.06
     Connect
    0.06
     orientation
    0.06
     lasts
    0.06
     lastname
    0.06
    Activations Density 0.001%

    No Known Activations