© 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. 22556
    Prev
    Next
    INDEX
    Explanations

    organizations or entities involved in legal or administrative contexts

    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
    465
    +0.12
    0.7%
    263
    +0.12
    0.7%
    190
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    499
    +0.12
    0.05
    397
    +0.12
    0.07
    331
    +0.12
    0.06
    Negative Logits
    :</
    -1.71
     Computing
    -1.48
    :(
    -1.46
     Namely
    -1.46
     Programming
    -1.44
     Their
    -1.43
     Note
    -1.41
     correctness
    -1.41
    >>
    -1.40
    valence
    -1.39
    POSITIVE LOGITS
    croft
    1.97
    ĨĴ
    1.79
     expire
    1.77
    %).
    1.65
    Ģ
    1.53
    .
    1.52
    leigh
    1.52
    ĥ½
    1.46
    mith
    1.44
    ake
    1.43
    Activations Density 0.696%

    No Known Activations