© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 12676
    Prev
    Next
    INDEX
    Explanations

    words related to the construction and arrangement of physical objects or components

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    348
    +0.19
    1.1%
    209
    +0.11
    0.6%
    208
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    304
    +0.19
    0.04
    309
    +0.11
    0.04
    5
    +0.10
    0.02
    Negative Logits
    ĥ½
    -3.67
    ¥
    -3.62
    ¼
    -3.53
    Į
    -3.52
    ľĵ
    -3.52
    Ī
    -3.48
    ı
    -3.46
    º
    -3.45
    Ĥ¬
    -3.44
    ĥ
    -3.44
    POSITIVE LOGITS
     thereon
    3.21
     thereto
    3.06
     therefrom
    2.79
     onto
    2.75
     therein
    2.74
     manually
    2.67
     laterally
    2.58
     vertically
    2.51
     radially
    2.48
     horizontally
    2.42
    Activations Density 0.253%

    No Known Activations