© 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. 30116
    Prev
    Next
    INDEX
    Explanations

    phrases indicating recommendations or endorsements

    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
    300
    +0.14
    0.8%
    264
    +0.13
    0.7%
    472
    +0.12
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    300
    +0.14
    0.04
    472
    +0.13
    0.04
    365
    +0.12
    0.02
    Negative Logits
     Islands
    -1.81
    utes
    -1.60
    theless
    -1.60
    bound
    -1.59
     Pradesh
    -1.58
    uous
    -1.55
    »¿
    -1.53
    elles
    -1.50
    iels
    -1.49
     sensed
    -1.46
    POSITIVE LOGITS
    ably
    1.77
     recommending
    1.76
    lication
    1.72
    fix
    1.65
     publication
    1.65
     recommend
    1.65
     recommendations
    1.58
     introdu
    1.55
     dosing
    1.48
     counseling
    1.47
    Activations Density 0.263%

    No Known Activations