© 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. 28306
    Prev
    Next
    INDEX
    Explanations

    the concept of crafting in various contexts

    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
    189
    +0.13
    0.7%
    274
    +0.12
    0.7%
    507
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    80
    +0.13
    0.01
    189
    +0.12
    0.01
    385
    +0.11
    0.01
    Negative Logits
     '</
    -2.41
    '?"
    -2.23
    >'
    -2.16
    ?'
    -2.16
    >';
    -2.10
    >`
    -2.01
     ',
    -1.89
     '_
    -1.86
     '
    -1.84
    "?"
    -1.81
    POSITIVE LOGITS
    enstein
    1.84
    bellum
    1.80
    illery
    1.79
    ruptcy
    1.76
    itect
    1.70
    falls
    1.70
    sburg
    1.67
    genstein
    1.64
    inux
    1.63
    ful
    1.62
    Activations Density 0.008%

    No Known Activations