© 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. 1-RES-SM
    11. 24252
    Prev
    Next
    INDEX
    Explanations

    instances of the words "the" and "was."

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    451
    +0.12
    0.6%
    453
    +0.11
    0.6%
    200
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    98
    +0.12
    0.00
    10
    +0.11
    0.00
    164
    +0.11
    0.00
    Negative Logits
    isans
    -1.69
    .]{}
    -1.62
    ees
    -1.55
    ership
    -1.55
    ulence
    -1.51
    İ
    -1.49
    ABILITY
    -1.45
    oise
    -1.45
    enberg
    -1.44
    ity
    -1.43
    POSITIVE LOGITS
    ipl
    1.46
    LL
    1.45
    awa
    1.43
    olved
    1.42
    PLC
    1.38
    olean
    1.38
    river
    1.35
    heres
    1.34
    ferenced
    1.34
     Limit
    1.33
    Activations Density 0.002%

    No Known Activations