© 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. 10686
    Prev
    Next
    INDEX
    Explanations

    CSS properties related to layout and spacing

    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
    348
    +0.27
    1.6%
    306
    +0.17
    1.0%
    307
    +0.17
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    348
    +0.27
    0.04
    129
    +0.17
    0.02
    7
    +0.17
    0.00
    Negative Logits
    play
    -1.55
    time
    -1.54
    frastructure
    -1.44
    mium
    -1.40
    ishing
    -1.38
    hib
    -1.36
     time
    -1.35
     administering
    -1.35
    dere
    -1.33
     performing
    -1.31
    POSITIVE LOGITS
    ī
    3.67
    ĻĤ
    3.50
    Į
    3.43
    ¬
    3.42
    į
    3.41
    Ĭ
    3.29
    Ļª
    3.28
    ļ
    3.26
    Ī
    3.23
    ¨
    3.21
    Activations Density 0.995%

    No Known Activations