© 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. 26834
    Prev
    Next
    INDEX
    Explanations

    technical jargon and terms related to programming concepts

    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
    17
    +0.23
    1.4%
    156
    +0.17
    1.0%
    379
    +0.16
    0.9%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    17
    +0.23
    -0.01
    240
    +0.17
    -0.08
    379
    +0.16
    0.28
    Negative Logits
    ↵↵↵
    -1.81
    ↵↵↵↵↵↵↵
    -1.60
    �
    -1.57
     unprecedented
    -1.57
    ↵↵↵↵
    -1.53
    ADVERTISEMENT
    -1.52
    ]>
    -1.52
    coat
    -1.51
    ICO
    -1.44
    ↵↵↵↵↵
    -1.44
    POSITIVE LOGITS
     etc
    2.00
     _
    1.81
     exceptions
    1.63
     liter
    1.58
     parentheses
    1.57
     {}
    1.54
     calculator
    1.54
     <-
    1.52
    ync
    1.50
     notation
    1.50
    Activations Density 15.264%

    No Known Activations