© 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. Residuals
    9. 4-RES-SM
    11. 10900
    Prev
    Next
    INDEX
    Explanations

    words and phrases related to support and funding initiatives

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    277
    +0.14
    0.8%
    293
    +0.12
    0.7%
    364
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    277
    +0.14
    0.02
    394
    +0.12
    0.03
    38
    +0.12
    0.03
    Negative Logits
    ish
    -2.28
    ishly
    -2.09
    ishes
    -1.80
    enburg
    -1.74
    elijk
    -1.73
    utz
    -1.61
    aves
    -1.59
    oning
    -1.54
     revenge
    -1.54
    eler
    -1.53
    POSITIVE LOGITS
     systems
    2.08
     mechanisms
    2.08
     vector
    2.04
     Vector
    2.02
     structures
    1.83
     functions
    1.82
     function
    1.73
     mechanism
    1.71
     roles
    1.69
     system
    1.69
    Activations Density 0.269%

    No Known Activations