© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW 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. 26367
    Prev
    Next
    INDEX
    Explanations

    references to dreams and visions

    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
    66
    +0.13
    0.7%
    505
    +0.13
    0.7%
    100
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    505
    +0.13
    0.04
    56
    +0.13
    -0.00
    100
    +0.12
    0.04
    Negative Logits
    chaft
    -1.62
    jours
    -1.52
     Spearman
    -1.50
     behalf
    -1.48
    Īĺ
    -1.47
    ĻĤ
    -1.42
    oppel
    -1.42
    «
    -1.41
    HECK
    -1.38
    ives
    -1.36
    POSITIVE LOGITS
    scape
    1.93
    ful
    1.87
    fest
    1.71
    t
    1.69
    sky
    1.66
    elijk
    1.61
    Works
    1.58
    iddle
    1.58
     worlds
    1.57
    orld
    1.57
    Activations Density 0.409%

    No Known Activations