© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    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. 2-RES-SM
    11. 28358
    Prev
    Next
    INDEX
    Explanations

    the word "originally" indicating the inception or beginning of an idea or concept

    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
    424
    +0.12
    0.7%
    23
    +0.11
    0.6%
    392
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    15
    +0.12
    0.02
    424
    +0.11
    0.01
    242
    +0.10
    0.02
    Negative Logits
    į
    -2.13
    Ĵ
    -2.13
    ľĵ
    -2.07
    ĥ
    -1.95
    Ħ
    -1.95
    ª
    -1.89
    Ĭ
    -1.86
    ees
    -1.85
    ļ
    -1.84
    º
    -1.82
    POSITIVE LOGITS
     extinct
    2.01
     Published
    1.42
     speaking
    1.40
     asleep
    1.39
     abandoned
    1.38
     underestimated
    1.35
     ration
    1.35
     PUBL
    1.34
    ya
    1.34
     outdated
    1.34
    Activations Density 0.117%

    No Known Activations