© 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. 4-RES-SM
    11. 12075
    Prev
    Next
    INDEX
    Explanations

    phrases that discuss various topics and research, focusing on the sharing of information and ideas

    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
    307
    +0.29
    1.7%
    306
    +0.18
    1.1%
    240
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    176
    +0.29
    0.12
    307
    +0.18
    0.13
    497
    +0.13
    0.03
    Negative Logits
    ĻĤ
    -2.48
    Ī
    -2.48
    Ĩ
    -2.46
    ¶
    -2.42
    ¡
    -2.40
    ħ
    -2.29
    °
    -2.19
    Ķ
    -2.19
    ´
    -2.13
    Īĺ
    -2.11
    POSITIVE LOGITS
     topics
    2.68
     classics
    2.30
     ideas
    2.25
     themes
    2.21
     trends
    2.17
     innovations
    2.15
     favorites
    2.13
     concepts
    2.13
     insights
    2.08
     genres
    2.08
    Activations Density 2.234%

    No Known Activations