© 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 EvalsExportsinterp-engineNEWAPI 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. 58967
    Prev
    Next
    INDEX
    Explanations

    common verbs

    np_max-act · gemini-2.0-flash

    The neuron reliably lights up on past‐tense auxiliary constructions (especially “had” or “was” plus a past participle), essentially flagging perfective narratives.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    stories that explore themes of self-exploration and boundary-pushing through unusual experiences.

    oai_token-act-pair · gpt-4o-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
     viêm
    -0.08
    ��
    -0.07
    lernen
    -0.07
    ệu
    -0.07
    grown
    -0.07
    رفة
    -0.06
    terr
    -0.06
    �
    -0.06
    archs
    -0.06
    ham
    -0.06
    POSITIVE LOGITS
     using
    0.07
    (sl
    0.06
    .Tool
    0.06
     handsome
    0.06
    Refer
    0.06
    {},
    0.06
     <",
    0.06
    Radi
    0.06
    ASCADE
    0.06
    [Double
    0.06
    Activations Density 0.086%

    No Known Activations