© 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 EvalsExportsinterp-engineNEWAPI 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. 57511
    Prev
    Next
    INDEX
    Explanations

    altering

    np_max-act · gemini-2.0-flash

    The neuron is looking for words that denote active security threats or malicious actions (e.g. “hacker,” “attacker,” “changing,” “manipulates”).

    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
    ourmet
    -0.07
    委
    -0.06
    autoreleasepool
    -0.06
    ロ
    -0.06
    _lo
    -0.06
    KeyDown
    -0.06
    .navigate
    -0.06
    (async
    -0.06
    iona
    -0.06
    INET
    -0.06
    POSITIVE LOGITS
    itemid
    0.08
    elere
    0.07
    하며
    0.07
     slav
    0.06
     habitual
    0.06
     Sergey
    0.06
     bif
    0.06
     ubiqu
    0.06
     transcend
    0.06
    (cmd
    0.06
    Activations Density 0.014%

    No Known Activations