© 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. 30354
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "Queen."

    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
    156
    +0.15
    0.8%
    488
    +0.13
    0.8%
    389
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    488
    +0.15
    0.02
    353
    +0.13
    0.02
    1
    +0.11
    0.02
    Negative Logits
    orters
    -1.84
    urred
    -1.69
    aught
    -1.69
    ached
    -1.68
    .'"
    -1.59
    illed
    -1.58
    orter
    -1.56
    ---|---
    -1.51
    itter
    -1.50
    ogn
    -1.46
    POSITIVE LOGITS
    doms
    1.89
    wave
    1.83
    zilla
    1.79
    holds
    1.68
    castle
    1.67
    mania
    1.60
    stown
    1.60
    fors
    1.58
    craft
    1.58
    mind
    1.58
    Activations Density 0.016%

    No Known Activations