© 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. 30265
    Prev
    Next
    INDEX
    Explanations

    terms related to mathematical or scientific classifications and dimensions

    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
    419
    +0.24
    1.4%
    436
    +0.17
    1.0%
    198
    +0.13
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    436
    +0.24
    0.06
    72
    +0.17
    0.05
    136
    +0.13
    -0.01
    Negative Logits
    Ļª
    -2.44
    ¢
    -2.25
    ¸
    -2.10
    ĸ´
    -1.98
    ¯
    -1.98
    Ļ
    -1.91
    Ĥ¬
    -1.90
    Ī
    -1.88
    »¿
    -1.88
    ¦
    -1.85
    POSITIVE LOGITS
    sciously
    1.70
    coated
    1.58
    ISHED
    1.44
    osto
    1.37
    plastic
    1.36
    posed
    1.33
     whom
    1.33
     tires
    1.32
     throws
    1.32
    eps
    1.32
    Activations Density 0.595%

    No Known Activations