Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    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. 40413
    Prev
    Next
    INDEX
    Explanations

    served fresh eaten

    np_max-act · gemini-2.0-flash

    The neuron remains nearly silent (all activations ≈0), indicating it does not respond to any consistent semantic or lexical pattern in these examples—it effectively detects nothing.

    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
    .grid
    -0.07
     flee
    -0.07
     outfield
    -0.06
     supporters
    -0.06
    .toCharArray
    -0.06
    (destination
    -0.06
    Trap
    -0.06
    .Undef
    -0.06
    Chicago
    -0.06
    "L
    -0.06
    POSITIVE LOGITS
     浙江
    0.07
     reacting
    0.06
    _site
    0.06
     angry
    0.06
    dol
    0.06
     prv
    0.06
    tolua
    0.06
     حالة
    0.06
     ge
    0.06
     Symphony
    0.06
    Activations Density 0.020%

    No Known Activations