© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 4-RES-SM
    11. 13219
    Prev
    Next
    INDEX
    Explanations

    HTML-like tags and their associated states or values

    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/4-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.4.hook_resid_post
    Hook Layer
    4
    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
    337
    +0.21
    1.4%
    307
    +0.18
    1.2%
    306
    +0.11
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    307
    +0.21
    0.14
    306
    +0.18
    0.05
    337
    +0.11
    0.09
    Negative Logits
    ©
    -4.10
    ī
    -3.94
    ľ
    -3.88
    §
    -3.88
    ¶
    -3.84
    ¥
    -3.81
    ·
    -3.80
    Ŀ
    -3.79
    ¤
    -3.77
    ł
    -3.76
    POSITIVE LOGITS
    xhtml
    1.75
    geries
    1.37
    schild
    1.26
    ellation
    1.19
     ablation
    1.14
     occult
    1.13
    xsd
    1.13
    imetry
    1.11
    lag
    1.09
    itage
    1.09
    Activations Density 2.911%

    No Known Activations