© 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. 10850
    Prev
    Next
    INDEX
    Explanations

    sections labeled as descriptions in various contexts

    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
    378
    +0.12
    0.7%
    301
    +0.12
    0.6%
    232
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    232
    +0.12
    0.03
    38
    +0.12
    0.01
    217
    +0.12
    0.02
    Negative Logits
    ĭ
    -2.28
    ĻĤ
    -2.26
    Ł
    -2.26
    Ĵ
    -2.08
    ı
    -2.07
    Īĺ
    -2.07
    ĩ
    -2.06
    Ľ
    -1.96
    ľ
    -1.95
    §
    -1.94
    POSITIVE LOGITS
    ality
    1.80
     follows
    1.79
    ably
    1.77
    igned
    1.73
    ally
    1.72
    aur
    1.69
    ogy
    1.68
    ongs
    1.56
    ately
    1.54
    :-
    1.53
    Activations Density 0.092%

    No Known Activations