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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 41547
    Prev
    Next
    INDEX
    Explanations

    is

    np_max-act · gemini-2.0-flash

    This neuron detects special structural tokens (e.g. the `<|start_header_id|>` and related marker tokens) used to mark headers or metadata in the conversation format.

    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
    .ax
    -0.06
     filmpjes
    -0.06
     damages
    -0.06
    ีข
    -0.06
     velice
    -0.06
     Forgotten
    -0.06
     AH
    -0.06
    jezd
    -0.06
    amaño
    -0.06
     πιο
    -0.06
    POSITIVE LOGITS
    alous
    0.07
    kanı
    0.06
     childbirth
    0.06
    }{$
    0.06
     asym
    0.06
     PATCH
    0.06
    _rt
    0.06
     bot
    0.06
    _allow
    0.06
    )tableView
    0.06
    Activations Density 0.015%

    No Known Activations