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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 91494
    Prev
    Next
    INDEX
    Explanations

    encoder

    np_max-act · gemini-2.0-flash

    The neuron fires on mentions of “autoencoder,” i.e. it detects occurrences of the word (and its parts) “autoencoder.”

    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
    atal
    -0.06
    _SENS
    -0.06
    盛
    -0.06
     완
    -0.06
     mobil
    -0.06
    Hyper
    -0.06
    Lic
    -0.06
    hwnd
    -0.06
    ATAL
    -0.06
    Camb
    -0.05
    POSITIVE LOGITS
    FLAGS
    0.07
     stress
    0.07
     stra
    0.06
     والتي
    0.06
    #error
    0.06
    conomics
    0.06
    ivan
    0.06
     Such
    0.06
     seperti
    0.06
     selfie
    0.06
    Activations Density 0.002%

    No Known Activations