© 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. 27112
    Prev
    Next
    INDEX
    Explanations

    references to programming languages and libraries, particularly related to data handling and APIs

    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
    231
    +0.15
    0.9%
    21
    +0.14
    0.9%
    233
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    21
    +0.15
    0.10
    107
    +0.14
    0.08
    422
    +0.13
    0.07
    Negative Logits
    iffs
    -1.54
    iques
    -1.52
    orph
    -1.48
    quisite
    -1.48
    ractions
    -1.47
    icals
    -1.45
    ique
    -1.42
    uckle
    -1.38
    vern
    -1.36
    itum
    -1.32
    POSITIVE LOGITS
     Âĵ
    1.29
     Walmart
    1.28
     fault
    1.27
    ²
    1.24
     Else
    1.22
     Starbucks
    1.22
     capacity
    1.20
    wikipedia
    1.20
     exceptions
    1.20
     Wal
    1.19
    Activations Density 4.423%

    No Known Activations