© 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. 9078
    Prev
    Next
    INDEX
    Explanations

    aspects related to effectiveness and performance metrics

    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
    307
    +0.27
    1.5%
    324
    +0.13
    0.8%
    490
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    324
    +0.27
    0.06
    291
    +0.13
    0.04
    215
    +0.12
    0.07
    Negative Logits
    unto
    -1.57
    plants
    -1.53
    ums
    -1.53
    etti
    -1.48
     expires
    -1.41
     reproduce
    -1.41
    ells
    -1.38
    sson
    -1.36
     completes
    -1.36
    iei
    -1.34
    POSITIVE LOGITS
     lack
    2.67
     inability
    2.51
     scarc
    2.37
     proximity
    2.17
     similarities
    2.13
    availability
    2.07
     simplicity
    2.02
     absence
    2.02
     inherent
    1.99
     accessibility
    1.96
    Activations Density 1.258%

    No Known Activations