Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 11985
    Prev
    Next
    INDEX
    Explanations

    technical terminology related to higher-dimensional theories in physics

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-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_resid_post
    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
    336
    +0.11
    0.6%
    501
    +0.11
    0.6%
    357
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    407
    +0.11
    0.06
    130
    +0.11
    0.08
    354
    +0.10
    0.05
    Negative Logits
    ĻĤ
    -1.89
     shipments
    -1.65
     Instagram
    -1.59
     Walmart
    -1.55
     USDA
    -1.49
     Owner
    -1.49
    owner
    -1.48
     shipment
    -1.48
     nutrients
    -1.48
     notifications
    -1.47
    POSITIVE LOGITS
    Theorem
    2.56
     theorem
    2.51
     Riemann
    2.41
     algebra
    2.40
     Schrödinger
    2.39
    theorem
    2.36
     cosmological
    2.32
     phenomenological
    2.31
     quantum
    2.31
     Theorem
    2.29
    Activations Density 1.670%

    No Known Activations