Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 28977
    Prev
    Next
    INDEX
    Explanations

    financial terms and concepts related to loans and collateral

    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
    156
    +0.28
    1.6%
    369
    +0.11
    0.7%
    17
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    455
    +0.28
    0.12
    287
    +0.11
    0.15
    127
    +0.11
    0.14
    Negative Logits
    Ŀ
    -2.65
    ŀ
    -2.62
    ¿½
    -2.62
    §
    -2.56
    ª
    -2.53
    ĻĤ
    -2.51
    ¦
    -2.35
    «
    -2.35
    ĸ
    -2.33
    ģ
    -2.32
    POSITIVE LOGITS
     replen
    1.45
     copying
    1.41
     centrifugation
    1.38
     digitally
    1.35
     manually
    1.34
     anhydr
    1.33
     Photoshop
    1.33
     subtracting
    1.31
     irradiation
    1.29
     delivery
    1.29
    Activations Density 3.944%

    No Known Activations