© 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
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 92367
    Prev
    Next
    INDEX
    Explanations

    underlying

    np_max-act · gemini-2.0-flash

    The neuron is tuned to German compound verb fragments (e.g. the “aut”/“geb” syllables in “gebaut”/“nachgebaut”) and related header‐ID tokens.

    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
    pra
    -0.07
     Fo
    -0.06
     Claw
    -0.06
    Successfully
    -0.06
     onlar
    -0.06
    -inv
    -0.06
     takdir
    -0.06
    alarda
    -0.06
    ца
    -0.06
     quand
    -0.06
    POSITIVE LOGITS
    uet
    0.08
    _Check
    0.07
     сила
    0.07
     sido
    0.07
     б
    0.06
    切
    0.06
    _thread
    0.06
     zoo
    0.06
    _System
    0.06
    _vel
    0.06
    Activations Density 0.507%

    No Known Activations