© 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. 1-RES-SM
    11. 29535
    Prev
    Next
    INDEX
    Explanations

    references to users or user-related entities

    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/1-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.1.hook_resid_post
    Hook Layer
    1
    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
    365
    +0.17
    1.0%
    26
    +0.11
    0.6%
    176
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    73
    +0.17
    0.02
    255
    +0.11
    0.02
    17
    +0.10
    0.01
    Negative Logits
    aped
    -1.73
    ¿½
    -1.68
    vt
    -1.59
    itely
    -1.56
    mpl
    -1.47
    §
    -1.46
    ĵ
    -1.43
    ctors
    -1.37
    fts
    -1.37
    ymers
    -1.34
    POSITIVE LOGITS
    helf
    1.91
     site
    1.65
     menu
    1.60
    pace
    1.58
    pective
    1.58
     table
    1.57
    CRIPT
    1.49
    urface
    1.44
     center
    1.44
    Controller
    1.43
    Activations Density 0.141%

    No Known Activations