© 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. 30801
    Prev
    Next
    INDEX
    Explanations

    coding syntax and structured data elements

    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
    161
    +0.15
    0.9%
    427
    +0.14
    0.8%
    419
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    111
    +0.15
    0.50
    156
    +0.14
    0.47
    71
    +0.13
    0.45
    Negative Logits
    ectomy
    -1.63
    inho
    -1.55
    osine
    -1.53
    ário
    -1.50
    Ignore
    -1.42
     goodbye
    -1.40
    ÅĻ
    -1.38
     equivalents
    -1.35
    že
    -1.35
     farewell
    -1.34
    POSITIVE LOGITS
    ľĵ
    1.96
    ¾
    1.89
    ĻĤ
    1.80
    IJ
    1.69
    ĥ
    1.62
    »
    1.60
    ĵ
    1.58
    ł
    1.52
    tern
    1.51
    ĩ
    1.46
    Activations Density 5.067%

    No Known Activations