© 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. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 30523
    Prev
    Next
    INDEX
    Explanations

    phrases and elements related to HTML or web design specifications

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    419
    +0.15
    0.9%
    17
    +0.13
    0.7%
    141
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    394
    +0.15
    0.14
    248
    +0.13
    0.12
    141
    +0.13
    0.09
    Negative Logits
    ĻĤ
    -3.34
    Ļª
    -2.76
    ¼
    -2.60
    ¿½
    -2.45
    ³
    -2.44
    Ļ
    -2.43
    Ĺ
    -2.38
    ·¸
    -2.33
    ĸ
    -2.27
    Ĭ
    -2.24
    POSITIVE LOGITS
     edition
    1.84
     /*!
    1.44
     hell
    1.43
    ``
    1.43
    ician
    1.36
     organization
    1.35
     prospective
    1.34
    :`
    1.34
    lege
    1.32
     ;
    1.32
    Activations Density 5.649%

    No Known Activations