Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 88345
    Prev
    Next
    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    The neuron consistently lights up on “technical” tokens – anything like file‐extensions, PHP/CodeIgniter identifiers or other acronyms/subwords that include dots or underscores rather than ordinary English words.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    content related to programming concepts, particularly focusing on coding examples and challenges.

    oai_token-act-pair · gpt-4o-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
    irty
    -0.08
    segment
    -0.06
    ifetime
    -0.06
    $string
    -0.06
     joking
    -0.06
     empowered
    -0.06
    	core
    -0.06
    unprocessable
    -0.06
    μος
    -0.06
     sesame
    -0.06
    POSITIVE LOGITS
     criticize
    0.06
     Buen
    0.06
    chai
    0.06
     Benchmark
    0.06
     broth
    0.06
    *b
    0.06
    oralType
    0.06
    .Bundle
    0.06
     shadow
    0.06
     ohne
    0.06
    Activations Density 1.365%

    No Known Activations