© 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. 2-RES-SM
    11. 13996
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions or notations involving parentheses

    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/2-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.2.hook_resid_post
    Hook Layer
    2
    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
    305
    +0.12
    0.7%
    187
    +0.11
    0.6%
    55
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    79
    +0.12
    0.01
    425
    +0.11
    0.01
    251
    +0.10
    0.01
    Negative Logits
    aked
    -1.65
    \$
    -1.57
    ~
    -1.40
    msgstr
    -1.33
     plus
    -1.32
    iph
    -1.31
    apshot
    -1.27
    IGNED
    -1.26
     thereto
    -1.26
     worldwide
    -1.24
    POSITIVE LOGITS
    bies
    1.46
    wright
    1.36
    ze
    1.35
    sen
    1.34
    anguage
    1.31
    smith
    1.30
     adj
    1.30
    rams
    1.28
    keit
    1.28
    gaard
    1.28
    Activations Density 0.021%

    No Known Activations