© 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
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 68790
    Prev
    Next
    INDEX
    Explanations

    confession

    np_max-act · gemini-2.0-flash

    The neuron activates on mentions of confessions or related admissions of guilt (e.g. “confession,” “confessions,” “voluntarily given”).

    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
    ур
    -0.08
    àu
    -0.07
    )new
    -0.07
    actic
    -0.07
     Onc
    -0.07
    uter
    -0.06
    ek
    -0.06
    hope
    -0.06
    (({
    -0.06
     Hog
    -0.06
    POSITIVE LOGITS
     Activity
    0.07
     система
    0.07
     RFC
    0.07
     carb
    0.07
     Declaration
    0.06
    -‐
    0.06
    _keywords
    0.06
     Ying
    0.06
     knees
    0.06
    _dec
    0.06
    Activations Density 0.007%

    No Known Activations