© 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 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. 57440
    Prev
    Next
    INDEX
    Explanations

    yet

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on words ending in the “-ed” past‐tense (or past‐participle) suffix.

    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
    GameObject
    -0.07
     회원
    -0.07
     funk
    -0.06
     trigger
    -0.06
    元素
    -0.06
     modeled
    -0.06
     dives
    -0.06
     dummy
    -0.06
     Correction
    -0.06
     attacked
    -0.06
    POSITIVE LOGITS
    createCommand
    0.06
    .fd
    0.06
    ,end
    0.06
     흐
    0.06
    ǐ
    0.06
     všechny
    0.06
    ButtonText
    0.06
    قي
    0.06
    ubl
    0.06
    _PP
    0.06
    Activations Density 0.146%

    No Known Activations