© 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. 28533
    Prev
    Next
    INDEX
    Explanations

    references to hyperlinks

    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
    256
    +0.17
    1.0%
    260
    +0.14
    0.8%
    472
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    260
    +0.17
    0.03
    256
    +0.14
    0.03
    329
    +0.13
    0.02
    Negative Logits
    Ļ
    -2.21
    ²
    -2.03
    ĥ½
    -1.87
    ¨
    -1.83
    »¿
    -1.83
    ľĵ
    -1.80
    ÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤÃĥÃĤ
    -1.73
    §
    -1.71
    <|outofrange|>
    -1.66
    -1.66
    POSITIVE LOGITS
    age
    2.21
    ages
    1.98
    up
    1.98
    points
    1.86
    marks
    1.83
    ings
    1.80
    back
    1.79
    hammer
    1.77
    aments
    1.77
    point
    1.75
    Activations Density 0.104%

    No Known Activations