© 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. 68902
    Prev
    Next
    INDEX
    Explanations

    expose

    np_max-act · gemini-2.0-flash

    The neuron flags terms expressing warnings of harm or security risks (e.g. “risk,” “expose,” “compromise,” “vulnerabilities”).

    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
     gez
    -0.07
    _SD
    -0.07
    _("
    -0.07
     IonicModule
    -0.07
     बस
    -0.06
    وك
    -0.06
    縮
    -0.06
    번호
    -0.06
     gute
    -0.06
     jezd
    -0.06
    POSITIVE LOGITS
     incumbent
    0.07
    �다
    0.07
     Foot
    0.07
    \Lib
    0.07
     segment
    0.06
     exposed
    0.06
    حي
    0.06
    HN
    0.06
     foot
    0.06
     PLAY
    0.06
    Activations Density 0.023%

    No Known Activations