© 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. 110193
    Prev
    Next
    INDEX
    Explanations

    Teen

    np_max-act · gemini-2.0-flash

    This neuron detects the numeric and single‐letter codes (object indices, action letters, and condition letters) and their surrounding punctuation when they appear in the patternized tuple output.

    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
    semicolon
    -0.07
    lining
    -0.06
     주요
    -0.06
    ewitness
    -0.06
     Zem
    -0.06
     MLP
    -0.06
    /twitter
    -0.06
     наук
    -0.06
    wcs
    -0.06
    _gb
    -0.06
    POSITIVE LOGITS
     fishermen
    0.06
     bush
    0.06
     beneficiary
    0.06
     Ground
    0.06
     다운받기
    0.06
     �
    0.06
     Ü
    0.06
     Ave
    0.06
    _img
    0.06
     blasts
    0.06
    Activations Density 0.016%

    No Known Activations