© 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. 4-RES-SM
    11. 13481
    Prev
    Next
    INDEX
    Explanations

    instances of the word "invitation" and its variations

    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
    21
    +0.11
    0.6%
    487
    +0.11
    0.6%
    219
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    199
    +0.11
    0.02
    501
    +0.11
    0.01
    181
    +0.11
    0.02
    Negative Logits
     footing
    -1.61
    llo
    -1.55
    herty
    -1.51
    sis
    -1.50
    pipe
    -1.49
    unnumbered
    -1.49
    board
    -1.47
     fault
    -1.46
    ered
    -1.43
    backs
    -1.43
    POSITIVE LOGITS
    Ĵ
    3.50
    ¦
    3.42
    IJ
    3.11
    ij
    3.07
    ¨
    3.05
    Ĺ
    3.04
    ĻĤ
    3.02
    º
    2.96
    ŀ
    2.94
    ¬
    2.85
    Activations Density 0.106%

    No Known Activations