© 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
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. MLP Post
    9. 0-MLP-SM
    11. 27030
    Prev
    Next
    INDEX
    Explanations

    references to social services and social welfare topics

    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
    362
    +0.16
    0.9%
    436
    +0.12
    0.7%
    503
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    362
    +0.16
    0.02
    436
    +0.12
    0.02
    503
    +0.11
    0.02
    Negative Logits
     deadly
    -1.67
     normal
    -1.66
     abusive
    -1.66
     certain
    -1.66
     compassionate
    -1.62
     paternal
    -1.62
     parental
    -1.61
     maternal
    -1.60
     sexual
    -1.59
     gentle
    -1.59
    POSITIVE LOGITS
    \$
    1.80
    ilage
    1.76
    hesis
    1.73
    IZE
    1.68
    ICES
    1.68
    bank
    1.65
    arman
    1.64
    RAM
    1.64
    Bank
    1.58
    {$
    1.58
    Activations Density 0.134%

    No Known Activations