© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 11986
    Prev
    Next
    INDEX
    Explanations

    instances of template code or syntax

    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
    78
    +0.17
    0.9%
    241
    +0.14
    0.8%
    27
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    27
    +0.17
    0.03
    241
    +0.14
    0.03
    78
    +0.11
    0.00
    Negative Logits
    onial
    -1.61
    inary
    -1.60
    itum
    -1.57
    worth
    -1.54
    ctic
    -1.48
    wise
    -1.47
    ubicin
    -1.46
    vity
    -1.43
     Laboratories
    -1.42
    tten
    -1.39
    POSITIVE LOGITS
    «
    3.47
    ©
    3.36
    Ķ
    3.04
    ľĵ
    3.03
    į
    3.00
    ĵ
    2.98
    ·¸
    2.96
    ¨
    2.95
    Ľ
    2.90
    §
    2.89
    Activations Density 0.188%

    No Known Activations