© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 30221
    Prev
    Next
    INDEX
    Explanations

    specific code constructs or programming terminology

    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
    153
    +0.17
    1.0%
    223
    +0.13
    0.7%
    79
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    74
    +0.17
    0.03
    315
    +0.13
    0.03
    185
    +0.13
    0.03
    Negative Logits
    .’”
    -1.50
    lections
    -1.48
    .'"
    -1.38
    verter
    -1.37
     neighboring
    -1.35
     generator
    -1.35
     neighbouring
    -1.33
     '%
    -1.33
     receivers
    -1.32
     cytoplasm
    -1.32
    POSITIVE LOGITS
    ĻĤ
    2.25
    ¬
    2.16
    ¾
    2.15
    ³
    2.04
    ¦
    1.92
    ŀ
    1.89
    §
    1.87
    ¿
    1.86
    ±
    1.82
    ij
    1.82
    Activations Density 0.238%

    No Known Activations