© 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. 12952
    Prev
    Next
    INDEX
    Explanations

    information related to academic qualifications, achievements, and contributions to respective fields

    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
    398
    +0.13
    0.8%
    466
    +0.13
    0.7%
    63
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    176
    +0.13
    0.06
    341
    +0.13
    0.03
    150
    +0.12
    0.05
    Negative Logits
     huh
    -1.75
     sense
    -1.52
    ption
    -1.40
    halt
    -1.35
     inevitable
    -1.34
     purpose
    -1.33
    wagen
    -1.30
     necessity
    -1.28
     simpler
    -1.26
     wrong
    -1.25
    POSITIVE LOGITS
    ĥ½
    2.73
    ª
    2.70
    ĥ
    2.57
    ½
    2.36
    ı
    2.32
    °
    2.28
    ¬
    2.25
    IJ
    2.24
    £
    2.24
    ¿
    2.22
    Activations Density 0.388%

    No Known Activations