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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports 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. 28758
    Prev
    Next
    INDEX
    Explanations

    Java function definitions and variable assignments

    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
    +0.13
    0.7%
    93
    +0.12
    0.7%
    294
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    157
    +0.13
    0.08
    22
    +0.12
    0.05
    170
    +0.11
    0.06
    Negative Logits
    edo
    -1.74
    amer
    -1.72
    [^
    -1.63
    grade
    -1.50
    versions
    -1.48
    )\]
    -1.47
    orf
    -1.46
    rfloor
    -1.46
    fixed
    -1.44
    opsis
    -1.43
    POSITIVE LOGITS
    ¥
    1.99
    Ĩ
    1.93
    ¿
    1.93
    ®
    1.86
     "#
    1.85
    ¾
    1.85
     "";
    1.83
     "@
    1.77
    ī
    1.77
    į
    1.76
    Activations Density 0.592%

    No Known Activations