© 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. 29171
    Prev
    Next
    INDEX
    Explanations

    expressions of enjoyment or pleasure 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/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
    23
    +0.13
    0.7%
    119
    +0.12
    0.7%
    376
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    231
    +0.13
    0.04
    119
    +0.12
    0.03
    112
    +0.12
    0.03
    Negative Logits
    OE
    -1.48
    UC
    -1.42
    SUM
    -1.40
    STE
    -1.39
    izumab
    -1.38
    enes
    -1.37
    oplastic
    -1.36
    STITUTE
    -1.36
    οÏħ
    -1.35
    Ïģί
    -1.34
    POSITIVE LOGITS
    cing
    1.88
    ful
    1.83
    ively
    1.78
     liking
    1.59
    ive
    1.57
    poons
    1.53
    fully
    1.48
     surprise
    1.47
     consistency
    1.47
     lay
    1.44
    Activations Density 0.217%

    No Known Activations