© 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. 28855
    Prev
    Next
    INDEX
    Explanations

    phrases related to statistical measures and comparisons in research 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
    156
    +0.24
    1.4%
    69
    +0.18
    1.1%
    478
    +0.17
    1.0%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    69
    +0.24
    0.20
    98
    +0.18
    -0.01
    478
    +0.17
    0.01
    Negative Logits
    erie
    -1.55
    away
    -1.55
    mina
    -1.54
    isor
    -1.54
    iets
    -1.50
    setminus
    -1.50
    esto
    -1.49
    intage
    -1.49
     palette
    -1.47
    ème
    -1.47
    POSITIVE LOGITS
    ¼
    2.56
    Ļª
    2.54
    ĻĤ
    2.51
    ¶
    2.43
    Ĭ
    2.33
    ¯
    2.30
    µ
    2.26
    ¸
    2.24
    ĥ
    2.21
    ½
    2.21
    Activations Density 1.387%

    No Known Activations