© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. MLP Post
    9. 0-MLP-SM
    11. 25512
    Prev
    Next
    INDEX
    Explanations

    repeated instances of the word "links"

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__mlp-sm_processed/0-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.0.hook_mlp_out
    Hook Layer
    0
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    376
    +0.19
    1.1%
    365
    +0.12
    0.7%
    329
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    329
    +0.19
    0.02
    143
    +0.12
    0.02
    178
    +0.11
    0.02
    Negative Logits
    rica
    -1.72
     lovely
    -1.54
     wonderful
    -1.53
     gente
    -1.50
     preponderance
    -1.49
    ureus
    -1.45
    woke
    -1.43
    cknowled
    -1.38
    oretical
    -1.36
     proud
    -1.34
    POSITIVE LOGITS
    mith
    1.97
    heet
    1.95
    helf
    1.91
    pot
    1.82
    hips
    1.82
    fel
    1.81
    hell
    1.81
    hop
    1.72
    hip
    1.68
    chool
    1.67
    Activations Density 0.019%

    No Known Activations