© 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. 27067
    Prev
    Next
    INDEX
    Explanations

    specific names or identifiers, particularly related to organizations or individuals

    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
    369
    +0.24
    1.4%
    156
    +0.19
    1.2%
    59
    +0.19
    1.1%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    410
    +0.24
    0.14
    59
    +0.19
    0.10
    369
    +0.19
    0.07
    Negative Logits
    izations
    -1.78
    izing
    -1.75
    cludes
    -1.68
    ization
    -1.62
    cluding
    -1.57
     favors
    -1.57
    ocument
    -1.55
    odies
    -1.54
    vdots
    -1.52
    isation
    -1.51
    POSITIVE LOGITS
    ĻĤ
    2.39
    ī
    2.35
    ±
    2.28
    ª
    2.25
    İ
    2.16
    ¹
    2.08
    Ĭ
    2.06
    ¿
    2.00
    ·
    1.92
    «
    1.91
    Activations Density 0.573%

    No Known Activations