© 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. 129650
    Prev
    Next
    INDEX
    Explanations

    code/math

    np_max-act · gemini-2.0-flash

    discussions about mathematical proofs and concepts.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron fires on the formal rule‐style notation of small‐step (structural) operational semantics—e.g. tokens like “structural,” “small‐step,” the LaTeX rule syntax (`⟨…⟩`, `→`), and program‐construct symbols such as `while`, `do`, `σ`, `B`, `S`.

    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
     Came
    -0.07
    сти
    -0.07
    CCI
    -0.06
     PlayStation
    -0.06
    иты
    -0.06
    ubby
    -0.06
    conscious
    -0.06
    スコ
    -0.06
    .Mask
    -0.06
    oteric
    -0.06
    POSITIVE LOGITS
     periodically
    0.07
     delaying
    0.06
     bekan
    0.06
     "$(
    0.06
     komplex
    0.06
    .barDockControl
    0.06
     "/"
    0.06
    (frames
    0.06
     Dangerous
    0.06
     langs
    0.06
    Activations Density 0.065%

    No Known Activations