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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 91650
    Prev
    Next
    INDEX
    Explanations

    This neuron primarily activates on placeholder entity tokens (e.g. “NAME,” “NAME_2,” etc.) rather than ordinary words.

    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
     outings
    -0.07
    PRS
    -0.07
     Pink
    -0.06
    بس
    -0.06
    Twenty
    -0.06
    �
    -0.06
     Gavin
    -0.06
    %).↵↵
    -0.06
     matchups
    -0.06
     peanuts
    -0.06
    POSITIVE LOGITS
     converter
    0.06
    _agent
    0.06
    .dll
    0.06
    _DATA
    0.06
    Undo
    0.06
     forgiveness
    0.06
     uvol
    0.06
     overrides
    0.06
     Cmd
    0.06
    uat
    0.06
    Activations Density 0.011%

    No Known Activations