© 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. 10720
    Prev
    Next
    INDEX
    Explanations

    terms related to object caching and performance metrics

    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
    111
    +0.24
    1.4%
    491
    +0.16
    0.9%
    11
    +0.14
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    298
    +0.24
    0.06
    11
    +0.16
    0.05
    504
    +0.14
    0.04
    Negative Logits
     “[
    -1.65
     Âĵ
    -1.64
    ]{};
    -1.57
    “[
    -1.50
     “
    -1.48
     âĢķ
    -1.44
    ]{}]{}
    -1.44
     denomin
    -1.42
     defendant
    -1.40
     "[
    -1.38
    POSITIVE LOGITS
    )|$(
    1.83
    ropolis
    1.82
    еÐ
    1.77
    heit
    1.73
     Age
    1.68
    Reviewer
    1.63
    anguage
    1.59
    riors
    1.58
    aceae
    1.56
     Against
    1.54
    Activations Density 0.286%

    No Known Activations