© 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. 14288
    Prev
    Next
    INDEX
    Explanations

    specific numerical and alphanumeric patterns

    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
    224
    +0.13
    0.7%
    136
    +0.12
    0.7%
    59
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    224
    +0.13
    0.04
    136
    +0.12
    0.00
    59
    +0.12
    0.04
    Negative Logits
     kidding
    -1.41
     following
    -1.41
     joking
    -1.38
    Tube
    -1.34
     MORE
    -1.34
     anonymity
    -1.34
     cheaper
    -1.33
     shortened
    -1.32
     apologize
    -1.31
     two
    -1.30
    POSITIVE LOGITS
    Ļª
    2.17
    IJ
    2.08
    ±
    1.97
    ¿
    1.97
    ¹
    1.96
    ¾
    1.91
    «
    1.86
    º
    1.84
    ŀ
    1.83
    °
    1.80
    Activations Density 0.365%

    No Known Activations