© 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. 2-MLP-SM
    11. 24332
    Prev
    Next
    INDEX
    Explanations

    statements of opinion and agreement

    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
    423
    +0.14
    0.8%
    499
    +0.14
    0.8%
    321
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    423
    +0.14
    0.18
    406
    +0.14
    0.14
    88
    +0.12
    0.12
    Negative Logits
    ITED
    -1.75
     suspicion
    -1.67
     adj
    -1.64
     remind
    -1.55
     disbelief
    -1.49
     nerve
    -1.47
     yell
    -1.46
    uss
    -1.45
     conjecture
    -1.45
     diag
    -1.44
    POSITIVE LOGITS
    pad
    1.63
    keeper
    1.59
     (£
    1.59
    arming
    1.59
    strip
    1.53
    enant
    1.52
    ophone
    1.46
    PES
    1.44
    ôle
    1.44
    agua
    1.43
    Activations Density 2.525%

    No Known Activations