© 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 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. 28616
    Prev
    Next
    INDEX
    Explanations

    instances of the word "agen."

    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
    385
    +0.16
    0.9%
    443
    +0.13
    0.7%
    40
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    385
    +0.16
    0.01
    420
    +0.13
    0.01
    80
    +0.10
    0.01
    Negative Logits
    "?"
    -2.07
    ="">
    -1.84
    ="
    -1.80
     ?>
    -1.70
    ?>
    -1.67
     another
    -1.67
     suspected
    -1.67
    ...?"
    -1.66
    else
    -1.63
    '?"
    -1.60
    POSITIVE LOGITS
    furt
    1.79
    omics
    1.58
    arium
    1.55
    borg
    1.51
    vre
    1.51
    gary
    1.49
    MAP
    1.48
    ics
    1.44
    bank
    1.42
    kers
    1.41
    Activations Density 0.009%

    No Known Activations