© 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. 25880
    Prev
    Next
    INDEX
    Explanations

    scientific terminology and concepts related to molecular structures and processes

    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
    156
    +0.20
    1.1%
    10
    +0.13
    0.7%
    340
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.20
    0.08
    271
    +0.13
    0.02
    156
    +0.13
    0.05
    Negative Logits
    wegian
    -1.70
    ascii
    -1.59
    iji
    -1.46
    quo
    -1.40
    ophyll
    -1.39
    ureus
    -1.38
     '\
    -1.36
     Navy
    -1.36
     swallow
    -1.34
     wrong
    -1.33
    POSITIVE LOGITS
    thirds
    1.93
    ķ
    1.87
    ement
    1.68
    eer
    1.63
    esis
    1.56
    emente
    1.54
    zilla
    1.50
    mente
    1.49
    arity
    1.47
    ISM
    1.47
    Activations Density 0.258%

    No Known Activations