© 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. Residuals
    9. 2-RES-SM
    11. 12636
    Prev
    Next
    INDEX
    Explanations

    instances of the word "in" at various frequencies

    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
    243
    +0.21
    1.2%
    282
    +0.16
    0.9%
    478
    +0.16
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    243
    +0.21
    0.10
    282
    +0.16
    0.07
    152
    +0.16
    0.05
    Negative Logits
    enderer
    -1.68
    ystems
    -1.54
    irs
    -1.52
     {#
    -1.48
    achine
    -1.48
    icular
    -1.39
     determinants
    -1.36
     angles
    -1.34
    amental
    -1.34
    ystem
    -1.30
    POSITIVE LOGITS
    ĥ½
    2.52
     addition
    2.47
    ↵
    2.46
    ↵
    2.46
     ↵
    2.46
                ↵
    2.46
    2.46
    2.46
    2.46
    ↵↵
    2.46
    Activations Density 0.212%

    No Known Activations