© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 28648
    Prev
    Next
    INDEX
    Explanations

    special characters or symbols, particularly those used in programming or encoding 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
    97
    +0.14
    0.8%
    427
    +0.12
    0.7%
    302
    +0.12
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    74
    +0.14
    0.07
    391
    +0.12
    0.08
    85
    +0.12
    0.05
    Negative Logits
    anas
    -1.59
    orman
    -1.50
    night
    -1.44
    called
    -1.43
    Station
    -1.43
    located
    -1.42
    porter
    -1.42
    helf
    -1.41
    culo
    -1.40
    ellers
    -1.40
    POSITIVE LOGITS
     regard
    2.00
     regards
    1.85
     Âĵ
    1.50
     respect
    1.44
     linger
    1.43
    ĻĤ
    1.42
     mold
    1.39
    OUN
    1.39
    itives
    1.37
    hood
    1.34
    Activations Density 2.614%

    No Known Activations