© 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. 31929
    Prev
    Next
    INDEX
    Explanations

    technical terms and version identifiers related to software or programming

    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
    136
    +0.15
    0.9%
    271
    +0.14
    0.8%
    122
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    271
    +0.15
    0.01
    136
    +0.14
    -0.02
    157
    +0.14
    0.13
    Negative Logits
     due
    -1.50
     defence
    -1.49
     defense
    -1.44
     means
    -1.40
     physic
    -1.40
     certainty
    -1.38
     numbered
    -1.36
    roads
    -1.33
     owing
    -1.33
     observation
    -1.33
    POSITIVE LOGITS
    ¿
    1.84
    1451450014514500
    1.77
    html
    1.74
    ³
    1.73
    ı
    1.67
    İ
    1.61
    º
    1.60
    Ļª
    1.54
    >&
    1.52
    аÐ
    1.52
    Activations Density 2.077%

    No Known Activations