© 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. 2-MLP-SM
    11. 24505
    Prev
    Next
    INDEX
    Explanations

    situations involving health risks and symptoms related to various conditions

    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/2-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.2.hook_mlp_out
    Hook Layer
    2
    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
    95
    +0.15
    0.8%
    412
    +0.14
    0.8%
    407
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    95
    +0.15
    0.14
    412
    +0.14
    0.15
    407
    +0.13
    0.14
    Negative Logits
    ³
    -2.39
     another
    -1.69
    ©
    -1.62
     Model
    -1.61
     Groups
    -1.40
     Severe
    -1.40
    osed
    -1.38
     Role
    -1.37
     Discussion
    -1.35
     Models
    -1.35
    POSITIVE LOGITS
    illa
    1.52
    ipl
    1.50
    gain
    1.49
     intervening
    1.48
    ”,
    1.47
    liga
    1.47
    ück
    1.44
     abroad
    1.43
    dem
    1.41
     won
    1.39
    Activations Density 3.272%

    No Known Activations