Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    © Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    3. Pythia-70M-Deduped
    4. Residuals
    5. 2-RES-SM
    6. 29653
    Prev
    Next
    INDEX
    Explanations

    references to libraries and collections in programming or documentation 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
    156
    +0.22
    1.3%
    451
    +0.11
    0.7%
    132
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    242
    +0.22
    0.02
    295
    +0.11
    0.02
    173
    +0.11
    0.01
    Negative Logits
    ĨĴ
    -4.02
    ĸ
    -3.91
    ¯
    -3.90
    Ĵ
    -3.67
    ³
    -3.64
    ĺ
    -3.50
    §
    -3.50
    »¿
    -3.39
    IJ
    -3.38
                                     
    -3.36
    POSITIVE LOGITS
    Dependencies
    2.00
    80211
    1.79
     system
    1.57
     structure
    1.55
    anguage
    1.50
    code
    1.49
     architecture
    1.49
    enium
    1.46
    forge
    1.46
    gebra
    1.43
    Activations Density 0.136%

    No Known Activations