© 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. 12635
    Prev
    Next
    INDEX
    Explanations

    HTML or XML tags and attributes

    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
    156
    +0.15
    0.8%
    439
    +0.14
    0.8%
    78
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    439
    +0.15
    0.04
    272
    +0.14
    0.04
    408
    +0.14
    0.03
    Negative Logits
     literature
    -1.54
     economies
    -1.49
     faster
    -1.49
     cheaper
    -1.46
     knowledge
    -1.46
     ability
    -1.43
     discretion
    -1.42
     theory
    -1.41
     abilities
    -1.41
     bargain
    -1.39
    POSITIVE LOGITS
    ĥ½
    3.31
    ½
    3.27
    ¢
    3.17
    ¥
    3.13
    ¦
    3.11
    «
    3.06
    ¬
    3.04
    ĭ
    3.02
    ²
    3.00
    ¤
    3.00
    Activations Density 0.191%

    No Known Activations