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. 55189
    Prev
    Next
    INDEX
    Explanations

    Historical texts

    np_max-act · gemini-2.0-flash

    The neuron specifically lights up on substantive, content‐bearing words in factual or explanatory passages—i.e. when the assistant is presenting detailed historical or informational content.

    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
    (IT
    -0.06
    "After
    -0.06
    DATES
    -0.06
    testimonial
    -0.06
     Sale
    -0.06
    _ts
    -0.06
    _CHANNELS
    -0.06
    gregate
    -0.06
    lated
    -0.06
     Kashmir
    -0.06
    POSITIVE LOGITS
    里面
    0.07
    npm
    0.07
    _BOOL
    0.07
     ←
    0.07
    _population
    0.07
     बन
    0.06
     kleinen
    0.06
    (columns
    0.06
    brakk
    0.06
    Autoresizing
    0.06
    Activations Density 0.105%

    No Known Activations