© 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. 30791
    Prev
    Next
    INDEX
    Explanations

    occurrences of the word "compounds."

    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
    102
    +0.13
    0.7%
    501
    +0.13
    0.7%
    141
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    258
    +0.13
    0.55
    56
    +0.13
    0.18
    203
    +0.12
    0.38
    Negative Logits
    mouth
    -1.76
     mail
    -1.61
    yzed
    -1.48
     cracks
    -1.48
    nix
    -1.47
     packages
    -1.46
    Äĩ
    -1.45
     courses
    -1.45
     course
    -1.43
     way
    -1.42
    POSITIVE LOGITS
    Ļ
    2.44
    Ĵ
    2.06
    ĵ
    2.01
    Ķ
    1.98
    ¢
    1.92
    ¾
    1.90
    ł
    1.90
    Ĺ
    1.80
    Ĩ
    1.76
    »
    1.72
    Activations Density 3.545%

    No Known Activations