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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. MLP Post
    5. 0-MLP-SM
    6. 26931
    Prev
    Next
    INDEX
    Explanations

    instances of the word "featured."

    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
    +0.22
    1.3%
    156
    +0.18
    1.0%
    148
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    372
    +0.22
    0.02
    228
    +0.18
    0.01
    411
    +0.14
    0.01
    Negative Logits
    roy
    -1.52
     Kyoto
    -1.49
    worth
    -1.35
    aned
    -1.32
    ivered
    -1.32
     fortune
    -1.32
     Acad
    -1.30
     Chancellor
    -1.29
    illian
    -1.29
    going
    -1.27
    POSITIVE LOGITS
     prominently
    1.56
     positions
    1.55
     areas
    1.41
    ently
    1.40
    positions
    1.40
    pieces
    1.40
     Areas
    1.37
    posing
    1.35
     area
    1.34
    pot
    1.33
    Activations Density 0.017%

    No Known Activations