© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 91954
    Prev
    Next
    INDEX
    Explanations

    Rapid increase

    np_max-act · gemini-2.0-flash

    The neuron activates on past-tense or past-participial forms ending in “-ed,” flagging words like “renowned,” “adored,” and “mortified.”

    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
    3
    -0.07
     �
    -0.07
    7
    -0.07
    optgroup
    -0.06
    5
    -0.06
     array
    -0.06
    7
    -0.06
    bove
    -0.06
     catchy
    -0.06
     versatile
    -0.06
    POSITIVE LOGITS
     soared
    0.07
    Origin
    0.07
    oland
    0.07
     Leafs
    0.07
     devour
    0.07
     hilarious
    0.06
    //---------------------------------------------------------------------------↵↵
    0.06
     pinpoint
    0.06
    Spawn
    0.06
     Got
    0.06
    Activations Density 0.084%

    No Known Activations