© 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. 26342
    Prev
    Next
    INDEX
    Explanations

    numeric values associated with dates or quantities

    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
    203
    +0.14
    0.8%
    210
    +0.12
    0.7%
    136
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    203
    +0.14
    0.09
    467
    +0.12
    0.06
    136
    +0.12
    -0.00
    Negative Logits
    ļ
    -3.05
    ĸ´
    -2.99
    Ń
    -2.95
    »
    -2.94
    ĥ½
    -2.93
    ¿½
    -2.88
    ¢
    -2.87
    ¶
    -2.85
    ģ
    -2.85
    ¡
    -2.83
    POSITIVE LOGITS
     Jr
    1.44
    000
    1.43
    etc
    1.37
     son
    1.35
     liking
    1.34
     etc
    1.33
     hooked
    1.31
    esque
    1.31
     both
    1.30
     downs
    1.29
    Activations Density 0.599%

    No Known Activations