© 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. 11239
    Prev
    Next
    INDEX
    Explanations

    instances of conditional and temporal phrases

    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
    487
    +0.13
    0.7%
    369
    +0.12
    0.7%
    243
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    243
    +0.13
    0.11
    465
    +0.12
    0.11
    274
    +0.11
    0.10
    Negative Logits
    ĻĤ
    -1.85
    Ļª
    -1.61
    âĪ¶
    -1.56
    ¥
    -1.51
    ¿½
    -1.50
    °
    -1.44
    IJ
    -1.41
    ľ
    -1.39
    §
    -1.36
    orum
    -1.33
    POSITIVE LOGITS
    ()).
    1.65
     wont
    1.57
     yours
    1.46
    $}
    1.44
     lasts
    1.44
    roz
    1.43
    $).
    1.41
     addict
    1.38
    letes
    1.38
    rue
    1.37
    Activations Density 1.056%

    No Known Activations