© 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. 11224
    Prev
    Next
    INDEX
    Explanations

    references to specific years or dates

    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
    308
    +0.13
    0.7%
    246
    +0.12
    0.7%
    399
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    51
    +0.13
    0.04
    308
    +0.12
    0.03
    364
    +0.12
    0.04
    Negative Logits
    ĻĤ
    -2.16
    «
    -1.65
    -1.64
    -1.63
    ↵
    -1.63
    -1.63
    ↵↵
    -1.63
    ↵
    -1.63
    ↵
    -1.63
    ↵↵↵
    -1.63
    POSITIVE LOGITS
    å¹´
    1.91
     anniversary
    1.83
    gio
    1.78
     rates
    1.56
    views
    1.54
    uses
    1.54
    rais
    1.53
     budget
    1.53
    's
    1.49
    morning
    1.48
    Activations Density 0.135%

    No Known Activations