© 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. 2-RES-SM
    11. 13969
    Prev
    Next
    INDEX
    Explanations

    references to clocks and time management concepts

    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
    113
    +0.17
    1.0%
    156
    +0.14
    0.8%
    110
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    113
    +0.17
    0.04
    110
    +0.14
    0.01
    148
    +0.13
    0.00
    Negative Logits
    lla
    -1.55
    fficiency
    -1.43
     financially
    -1.38
    llo
    -1.37
     circum
    -1.36
    fficient
    -1.35
     situ
    -1.33
     respectful
    -1.32
     prospective
    -1.32
     manageable
    -1.31
    POSITIVE LOGITS
    astic
    1.82
    hetti
    1.71
    orian
    1.70
    over
    1.60
    iness
    1.60
    avia
    1.59
     intervals
    1.57
    shake
    1.57
    well
    1.56
     oscillation
    1.56
    Activations Density 0.256%

    No Known Activations