© 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. 1-RES-SM
    11. 25905
    Prev
    Next
    INDEX
    Explanations

    references to Python programming language syntax and elements

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    317
    +0.21
    1.2%
    369
    +0.14
    0.8%
    134
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    317
    +0.21
    0.02
    230
    +0.14
    0.02
    139
    +0.12
    0.02
    Negative Logits
    Ĭ
    -1.68
    Ļª
    -1.49
    MHz
    -1.47
    hip
    -1.46
    itance
    -1.44
     brow
    -1.42
    ighed
    -1.41
    igh
    -1.40
     MERCHANTABILITY
    -1.36
    itus
    -1.36
    POSITIVE LOGITS
    erals
    1.82
    ricular
    1.70
    rote
    1.63
    lectual
    1.61
     prominently
    1.61
     filings
    1.58
    gments
    1.57
    rano
    1.54
    ruvate
    1.54
    :`
    1.47
    Activations Density 0.078%

    No Known Activations