© 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
    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. 42758
    Prev
    Next
    INDEX
    Explanations

    colon

    np_max-act · gemini-2.0-flash

    The neuron detects the colon character used to separate metadata field names (e.g. “author:”, “date:”, “title:”) from their values in document front matter.

    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
    713
    -0.07
    _MARK
    -0.06
    urrect
    -0.06
     TP
    -0.06
    ман
    -0.06
     тебе
    -0.06
     Murder
    -0.06
    .AddParameter
    -0.06
     Blocked
    -0.06
     Ottawa
    -0.06
    POSITIVE LOGITS
    (GTK
    0.07
    seconds
    0.07
    inctions
    0.06
    uhn
    0.06
     broadly
    0.06
    emaker
    0.06
    TECT
    0.06
     Rich
    0.06
     �
    0.06
     $_
    0.06
    Activations Density 0.000%

    No Known Activations