© 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. 25844
    Prev
    Next
    INDEX
    Explanations

    instances of the "ng" suffix in words

    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
    +0.18
    1.1%
    59
    +0.15
    0.9%
    376
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    59
    +0.18
    0.01
    259
    +0.15
    0.01
    344
    +0.12
    0.01
    Negative Logits
    iginally
    -1.56
    EOF
    -1.51
    es
    -1.45
    Introduction
    -1.42
    lament
    -1.39
     cref
    -1.38
    endez
    -1.38
    omitempty
    -1.37
     "*
    -1.36
    soever
    -1.36
    POSITIVE LOGITS
    ¹
    2.85
    ĸ
    2.83
    Ŀ
    2.83
    Ħ
    2.78
    §
    2.73
    ª
    2.71
    ļ
    2.71
    ®
    2.69
    ±
    2.69
    ģ
    2.68
    Activations Density 0.043%

    No Known Activations