© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 7558
    Prev
    Next
    INDEX
    Explanations

    has

    np_max-act · gemini-2.0-flash

    The neuron tends to pick out common topical nouns (e.g. “landscape,” “ministry,” “business,” “library”) and related quantifiers (“all,” “every,” “its”) that signal the main subjects or themes of a passage.

    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
    """
    -0.07
     Origins
    -0.06
    Moh
    -0.06
    onestly
    -0.06
    <|eot_id|>
    -0.06
    BJ
    -0.06
    교회
    -0.06
    .custom
    -0.06
    .sky
    -0.06
    ********************
    -0.05
    POSITIVE LOGITS
     xAxis
    0.07
     lookahead
    0.07
    γκε
    0.07
    องท
    0.07
    [port
    0.07
    سبب
    0.07
    [DllImport
    0.06
     pornstar
    0.06
    latable
    0.06
     багат
    0.06
    Activations Density 0.041%

    No Known Activations