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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 38287
    Prev
    Next
    INDEX
    Explanations

    open source

    np_max-act · gemini-2.0-flash

    The neuron activates on words expressing planned or potential actions and benefits—modal/future‐tense verbs and related terms used in persuasive or justificatory statements (e.g. “will,” “allow,” “benefits,” “bring”).

    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
     straightforward
    -0.07
    isan
    -0.07
    视
    -0.06
    期待
    -0.06
    Rs
    -0.06
     Men
    -0.06
     frost
    -0.06
    -independent
    -0.06
    590
    -0.06
     fractures
    -0.06
    POSITIVE LOGITS
    bih
    0.07
    ungkin
    0.07
     zvý
    0.07
     війсь
    0.06
    -basket
    0.06
    .share
    0.06
    생활
    0.06
    <\/
    0.06
     pageIndex
    0.06
    _filenames
    0.06
    Activations Density 0.024%

    No Known Activations