© 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. 57539
    Prev
    Next
    INDEX
    Explanations

    not

    np_max-act · gemini-2.0-flash

    This neuron responds to negative or subjective expressions—particularly negations (“not,” “doesn’t”) and negative/opinion words (e.g. “horrible,” “want,” “need”).

    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
    ヘ
    -0.08
    сп
    -0.07
     Unsupported
    -0.07
    .but
    -0.07
    Kernel
    -0.06
     plano
    -0.06
     Het
    -0.06
    }s
    -0.06
    resh
    -0.06
    maktan
    -0.06
    POSITIVE LOGITS
    	ast
    0.07
    permission
    0.07
    achusetts
    0.06
    습니다
    0.06
    ☆
    0.06
    です
    0.06
    �니다
    0.06
     Physiology
    0.06
     خودش
    0.06
     underage
    0.06
    Activations Density 0.030%

    No Known Activations