© 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. 0-MLP-SM
    11. 25100
    Prev
    Next
    INDEX
    Explanations

    instances of the word "works" and its variations in various 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/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
    Neuron Alignment
    Index
    Value
    % of L₁
    144
    +0.16
    0.9%
    260
    +0.13
    0.7%
    376
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    144
    +0.16
    0.03
    347
    +0.13
    0.03
    260
    +0.11
    0.03
    Negative Logits
    apine
    -1.50
     boot
    -1.50
     hopes
    -1.49
     privileged
    -1.44
     welcome
    -1.43
     feared
    -1.39
     wealthy
    -1.39
     patriarch
    -1.35
     grateful
    -1.35
    verted
    -1.33
    POSITIVE LOGITS
    manship
    2.51
    icles
    2.30
    piece
    2.25
    pieces
    2.18
    paces
    2.12
    bench
    2.08
    icle
    2.08
    hops
    2.02
    pace
    1.95
    place
    1.93
    Activations Density 0.016%

    No Known Activations