© 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. MLP Post
    9. 2-MLP-SM
    11. 22538
    Prev
    Next
    INDEX
    Explanations

    phrases that indicate categories or classifications

    oai_token-act-pair · gpt-4o-mini
    Triggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/2-mlp-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_mlp_out
    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
    64
    +0.16
    0.9%
    83
    +0.14
    0.8%
    382
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    64
    +0.16
    0.04
    295
    +0.14
    0.04
    207
    +0.12
    0.03
    Negative Logits
    HP
    -1.42
     Tibetan
    -1.37
    CDATA
    -1.37
    STEM
    -1.34
    IPV
    -1.30
     marginal
    -1.27
    ubern
    -1.26
    ICC
    -1.24
    ISP
    -1.24
    Viet
    -1.23
    POSITIVE LOGITS
    İ
    2.10
    ´
    2.03
    ł
    2.02
    Ń
    1.98
    º
    1.81
    ħ
    1.75
    ½
    1.73
    ¸
    1.71
    ķ
    1.68
    ij
    1.66
    Activations Density 0.075%

    No Known Activations