© 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. 11243
    Prev
    Next
    INDEX
    Explanations

    instances of high-level descriptive phrases or adjectives that indicate quality or importance

    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
    203
    +0.14
    0.8%
    397
    +0.12
    0.7%
    130
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    203
    +0.14
    0.09
    147
    +0.12
    0.08
    397
    +0.11
    0.07
    Negative Logits
    wagen
    -1.93
    orr
    -1.79
    ilde
    -1.67
    ota
    -1.65
    ania
    -1.55
    ahoma
    -1.54
    ibase
    -1.51
    osin
    -1.50
    adesh
    -1.49
    osi
    -1.48
    POSITIVE LOGITS
    yet
    1.66
     yet
    1.60
     but
    1.50
    spring
    1.48
    biased
    1.45
     whereby
    1.43
     born
    1.41
    chemical
    1.37
     especially
    1.37
     albeit
    1.36
    Activations Density 0.714%

    No Known Activations