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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Qwen2.5-7B-IT
    4. Resid Post - 131k
    5. 23-RESID-POST-AA
    6. 130927
    Prev
    Next
    INDEX
    Explanations

    Website text

    np_max-act-logits · gemini-2.0-flashTriggered by @majianfei030706
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-qwen2.5-7b-instruct/resid_post_layer_23/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.23.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     abolished
    -0.08
    öt
    -0.08
    粞
    -0.07
    Accordion
    -0.06
     Schwartz
    -0.06
     smashed
    -0.06
    >(()
    -0.06
    dit
    -0.06
     masa
    -0.06
     compromising
    -0.06
    POSITIVE LOGITS
    �
    0.08
    ["@
    0.07
    :].
    0.07
     Był
    0.07
    쾀
    0.07
    /tiny
    0.07
    \S
    0.06
    شراء
    0.06
    '].
    0.06
    ŷ
    0.06
    Activations Density 0.157%

    No Known Activations