© 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 EvalsExportsAPI 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. 29290
    Prev
    Next
    INDEX
    Explanations

    references to specific locations or landmarks associated with the word "Hill."

    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
    369
    +0.14
    0.8%
    363
    +0.11
    0.6%
    166
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    31
    +0.14
    0.02
    94
    +0.11
    0.02
    363
    +0.11
    0.02
    Negative Logits
    usterity
    -2.03
    ĥ½
    -1.85
    ities
    -1.79
    ging
    -1.62
     whole
    -1.58
    ¯
    -1.53
     fatty
    -1.51
     white
    -1.51
    ellees
    -1.49
    iling
    -1.48
    POSITIVE LOGITS
    hurst
    1.92
    craft
    1.87
    bourne
    1.86
    yard
    1.79
     needs
    1.71
    emos
    1.69
    iewicz
    1.67
    ema
    1.67
    forge
    1.64
     creates
    1.63
    Activations Density 0.013%

    No Known Activations