© 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. 2-RES-SM
    11. 10673
    Prev
    Next
    INDEX
    Explanations

    critical details and significant events in narratives

    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
    422
    +0.17
    1.0%
    95
    +0.13
    0.8%
    12
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    95
    +0.17
    0.12
    407
    +0.13
    0.12
    412
    +0.13
    0.16
    Negative Logits
    IJ
    -2.84
    «
    -2.65
    ĺ
    -2.51
    ľ
    -2.48
    §
    -2.48
    ĻĤ
    -2.46
    ©
    -2.45
    ¥
    -2.44
    °
    -2.42
    ħ
    -2.42
    POSITIVE LOGITS
     doors
    1.58
    boards
    1.57
     decks
    1.47
    doors
    1.45
    aper
    1.41
     parl
    1.36
    GET
    1.35
     Facility
    1.35
     Rd
    1.34
    Tools
    1.33
    Activations Density 5.584%

    No Known Activations