© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 7804
    Prev
    Next
    INDEX
    Explanations

    reasons/reason

    np_max-act · gemini-2.0-flash

    This neuron detects language expressing emotional reactions or concerns—especially words about worry, fear, surprise, or shifts in perception.

    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
     divisive
    -0.07
    _index
    -0.06
     strconv
    -0.06
     thôi
    -0.06
    _DOC
    -0.06
     selecting
    -0.06
     komb
    -0.06
     carve
    -0.06
     linker
    -0.06
    .writeInt
    -0.06
    POSITIVE LOGITS
    ?><
    0.08
     opi
    0.07
     Karachi
    0.06
    AdapterManager
    0.06
     BAD
    0.06
    (EIF
    0.06
    ]";↵
    0.06
     }}"></
    0.06
    े।↵
    0.06
     BALL
    0.06
    Activations Density 0.112%

    No Known Activations