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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 1-RES-SM
    6. 26714
    Prev
    Next
    INDEX
    Explanations

    quantitative data and numerical references related to measurements or statistics

    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
    477
    +0.13
    0.7%
    23
    +0.13
    0.7%
    511
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    477
    +0.13
    0.12
    458
    +0.13
    0.09
    307
    +0.12
    0.10
    Negative Logits
    eer
    -1.52
     genre
    -1.47
    scheme
    -1.42
     waived
    -1.40
    OSS
    -1.38
    ledge
    -1.38
     \-
    -1.37
    materials
    -1.34
    Competing
    -1.33
     maker
    -1.32
    POSITIVE LOGITS
    th
    1.73
    uous
    1.56
    fold
    1.52
    ured
    1.48
     surveyed
    1.47
    ermost
    1.47
     th
    1.47
     fractions
    1.46
    ugin
    1.44
    aska
    1.44
    Activations Density 0.476%

    No Known Activations