© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 14480
    Prev
    Next
    INDEX
    Explanations

    references to data and information within various contexts

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    354
    +0.15
    0.8%
    431
    +0.13
    0.7%
    150
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    354
    +0.15
    0.04
    431
    +0.13
    0.04
    371
    +0.13
    0.03
    Negative Logits
    Ļª
    -3.96
    ĵ
    -3.80
    ĨĴ
    -3.75
    ¦
    -3.71
    ¯
    -3.69
    §
    -3.67
    į
    -3.61
    ł
    -3.55
    IJ
    -3.49
    ³
    -3.39
    POSITIVE LOGITS
    flow
    2.39
     collection
    2.34
     mining
    2.25
     analytics
    2.18
    frame
    2.15
    frames
    2.09
    share
    2.03
     storage
    2.01
     acquisition
    1.99
     ware
    1.99
    Activations Density 0.103%

    No Known Activations