© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 14224
    Prev
    Next
    INDEX
    Explanations

    numerical expressions and mathematical operations

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Neuron Alignment
    Index
    Value
    % of L₁
    317
    +0.14
    0.8%
    392
    +0.13
    0.7%
    331
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    225
    +0.14
    0.03
    392
    +0.13
    0.03
    317
    +0.11
    0.03
    Negative Logits
    ortality
    -1.58
    ĻĤ
    -1.56
    ĸ´
    -1.46
    nexpected
    -1.46
    ķ
    -1.45
     onward
    -1.44
     transferred
    -1.44
    ders
    -1.40
    swick
    -1.38
    noreply
    -1.36
    POSITIVE LOGITS
     =>
    1.65
     trap
    1.49
     modifier
    1.48
     shorth
    1.39
    1.36
     stride
    1.31
    Setter
    1.28
    plus
    1.27
    Scope
    1.27
    legate
    1.23
    Activations Density 0.191%

    No Known Activations