© 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
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 28252
    Prev
    Next
    INDEX
    Explanations

    superlative forms of adjectives

    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
    237
    +0.16
    0.9%
    269
    +0.12
    0.6%
    411
    +0.10
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    237
    +0.16
    0.03
    58
    +0.12
    0.02
    67
    +0.10
    0.01
    Negative Logits
    ´
    -2.53
    º
    -2.42
    µ
    -2.38
    ¶
    -2.36
    ĵ
    -2.34
    ĸ
    -2.33
    ©
    -2.32
    ¿½
    -2.31
    ½
    -2.30
    «
    -2.29
    POSITIVE LOGITS
    aurant
    2.14
    osterone
    1.93
    imony
    1.90
    ivals
    1.85
    oppel
    1.83
    ibular
    1.79
    ruct
    1.63
    rated
    1.57
    anding
    1.54
    robe
    1.53
    Activations Density 0.122%

    No Known Activations