© 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. 10895
    Prev
    Next
    INDEX
    Explanations

    terms and phrases related to spatial orientation and alignment

    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
    419
    +0.13
    0.7%
    325
    +0.12
    0.7%
    110
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    325
    +0.13
    0.03
    137
    +0.12
    0.03
    18
    +0.11
    0.02
    Negative Logits
    rooms
    -1.89
    urop
    -1.85
    ting
    -1.84
    aceut
    -1.84
    roy
    -1.78
    riev
    -1.78
    derer
    -1.75
    angered
    -1.69
    chers
    -1.67
    glich
    -1.67
    POSITIVE LOGITS
     direction
    2.41
    isation
    2.19
    ities
    2.07
     scrolling
    2.06
     gradients
    2.04
     alignment
    2.03
    ized
    2.03
     distribution
    2.03
     axis
    2.02
    ity
    2.00
    Activations Density 0.136%

    No Known Activations