© 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. 57321
    Prev
    Next
    INDEX
    Explanations

    cord

    np_max-act · gemini-2.0-flash

    This neuron detects the subword “cord,” activating on occurrences of “cord” (e.g. in “cord blood”) or words containing that substring (e.g. “cordless”).

    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
    astle
    -0.08
    npj
    -0.07
     Asheville
    -0.07
     Bay
    -0.07
    ALLE
    -0.07
     fallen
    -0.07
    vrier
    -0.07
    wei
    -0.07
     neuen
    -0.07
    oklyn
    -0.06
    POSITIVE LOGITS
     cord
    0.17
     Cord
    0.13
     cords
    0.13
    cordova
    0.10
     chord
    0.09
    contained
    0.08
    cord
    0.08
     crib
    0.08
    ird
    0.08
     discord
    0.08
    Activations Density 0.006%

    No Known Activations