© 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. 31010
    Prev
    Next
    INDEX
    Explanations

    Negative sentiment

    np_max-act · gemini-2.0-flash

    themes related to justice and reformation.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron selectively activates on words invoking conflict or its resolution—terms like “war,” “violence,” “peace,” and “enemies.”

    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
     committed
    -0.07
    azzo
    -0.07
     Damon
    -0.07
     Manhattan
    -0.07
     instagram
    -0.06
    Stan
    -0.06
    We
    -0.06
    amam
    -0.06
    εδ
    -0.06
     confiscated
    -0.06
    POSITIVE LOGITS
    _gui
    0.07
    ragen
    0.07
    /board
    0.07
    ricanes
    0.06
     Jurassic
    0.06
     bisher
    0.06
    ,\"
    0.06
    овал
    0.06
    รายการ
    0.06
    /ion
    0.06
    Activations Density 0.061%

    No Known Activations