© 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. 9279
    Prev
    Next
    INDEX
    Explanations

    punctuation marks, particularly periods

    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
    +0.45
    3.2%
    255
    +0.13
    1.0%
    78
    +0.11
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    21
    +0.45
    0.05
    255
    +0.13
    0.01
    450
    +0.11
    0.02
    Negative Logits
    written
    -1.46
    bing
    -1.38
    bed
    -1.35
    etting
    -1.29
    iating
    -1.20
     mismatch
    -1.20
    pered
    -1.17
     reception
    -1.16
    imization
    -1.11
    das
    -1.10
    POSITIVE LOGITS
    S
    1.63
    rsquo
    1.43
    Qaeda
    1.30
     pylori
    1.21
     stato
    1.21
     coli
    1.17
    ActivityThread
    1.17
     philosoph
    1.16
     sins
    1.16
     heels
    1.16
    Activations Density 0.083%

    No Known Activations