© 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. 2-RES-SM
    11. 10227
    Prev
    Next
    INDEX
    Explanations

    technical terms related to audio and electromagnetic systems

    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
    507
    +0.22
    1.3%
    47
    +0.15
    0.9%
    459
    +0.15
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    78
    +0.22
    0.06
    507
    +0.15
    0.05
    271
    +0.15
    -0.02
    Negative Logits
     matern
    -1.58
     Acad
    -1.54
     County
    -1.50
     bye
    -1.49
    eligible
    -1.44
     nationals
    -1.43
     citizenship
    -1.43
     Borough
    -1.40
     ripe
    -1.38
     Greek
    -1.38
    POSITIVE LOGITS
    º
    3.13
    Ł
    2.81
    Ĥ
    2.80
    Ļ
    2.79
    ĸ
    2.79
    ľ
    2.77
    ®
    2.75
    Ģ
    2.75
    ¼
    2.72
    Ĩ
    2.70
    Activations Density 1.688%

    No Known Activations