© 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. 56901
    Prev
    Next
    INDEX
    Explanations

    other

    np_max-act · gemini-2.0-flash

    The neuron responds to connective phrasing that links or compares two things (e.g. “both … and …,” “other,” “two processes”), marking sentences that draw parallels or relationships.

    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
     tjejer
    -0.08
    Alarm
    -0.07
    principal
    -0.07
    .offsetHeight
    -0.07
    Intent
    -0.06
    kj
    -0.06
    _policy
    -0.06
     Lodge
    -0.06
     Alarm
    -0.06
    _evaluation
    -0.06
    POSITIVE LOGITS
    (cols
    0.06
     सभ
    0.06
     dos
    0.06
     bif
    0.06
     Bravo
    0.06
     deep
    0.06
    hr
    0.06
     ear
    0.06
     wh
    0.06
     był
    0.06
    Activations Density 0.025%

    No Known Activations