Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 130716
    Prev
    Next
    INDEX
    Explanations

    ova

    np_max-act · gemini-2.0-flash

    The neuron activates on mentions of “Cordova,” i.e. when the text refers to the Cordova framework or its plugins.

    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
    secutive
    -0.07
     recurrent
    -0.07
    (UnityEngine
    -0.07
     recurs
    -0.07
    ICIENT
    -0.06
    COVERY
    -0.06
     autoplay
    -0.06
    datatype
    -0.06
     observer
    -0.06
    laus
    -0.06
    POSITIVE LOGITS
    \htdocs
    0.08
     charm
    0.07
     revel
    0.07
     TA
    0.07
    arrass
    0.07
    ova
    0.06
    -www
    0.06
     mashed
    0.06
    Truth
    0.06
    =\'
    0.06
    Activations Density 0.002%

    No Known Activations