© 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. 28202
    Prev
    Next
    INDEX
    Explanations

    instances of the word "to" in various contexts

    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
    478
    +0.21
    1.2%
    69
    +0.16
    0.9%
    47
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.21
    0.07
    505
    +0.16
    0.05
    434
    +0.12
    0.05
    Negative Logits
    ishly
    -2.00
     measures
    -1.72
    anth
    -1.56
    ively
    -1.51
    ly
    -1.50
    icul
    -1.49
    uous
    -1.43
    ally
    -1.40
    Description
    -1.39
     factors
    -1.38
    POSITIVE LOGITS
     '/
    1.88
    gg
    1.70
     replace
    1.67
     favor
    1.65
    fu
    1.65
     recreate
    1.64
     reproduce
    1.63
     avoid
    1.60
    asts
    1.60
     tmp
    1.54
    Activations Density 0.226%

    No Known Activations