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

    Neuronpedia

    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. 5992
    Prev
    Next
    INDEX
    Explanations

    fighting

    np_max-act · gemini-2.0-flash

    The neuron activates on words and phrases that denote combating, reducing, or eliminating social ills (e.g., “combat,” “eliminate,” “reduce,” “end the cycle”).

    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
    看到
    -0.07
     druž
    -0.06
     그러
    -0.06
    ære
    -0.06
     mlad
    -0.06
    Invalid
    -0.06
     müm
    -0.06
    _"+
    -0.06
     pressures
    -0.06
    _TYPED
    -0.06
    POSITIVE LOGITS
     combating
    0.07
    End
    0.07
    (baseUrl
    0.07
    ств
    0.06
     prohibiting
    0.06
     Elim
    0.06
    pectives
    0.06
    箱
    0.06
    Obama
    0.06
     filib
    0.06
    Activations Density 0.021%

    No Known Activations