© 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. 4-RES-SM
    11. 13334
    Prev
    Next
    INDEX
    Explanations

    references to the passage of time or end points in situations

    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
    466
    +0.17
    1.0%
    307
    +0.16
    0.9%
    68
    +0.09
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    106
    +0.17
    0.03
    485
    +0.16
    0.03
    69
    +0.09
    0.05
    Negative Logits
    men
    -1.63
    Bg
    -1.46
    esses
    -1.42
    aders
    -1.42
    ![
    -1.38
    /+
    -1.37
    UTF
    -1.36
    \(
    -1.34
    References
    -1.34
    guards
    -1.34
    POSITIVE LOGITS
     gestation
    2.00
     2016
    1.98
     October
    1.96
     2015
    1.94
     2017
    1.93
     July
    1.92
     each
    1.92
     November
    1.90
     September
    1.90
     June
    1.87
    Activations Density 0.382%

    No Known Activations