© 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
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 28680
    Prev
    Next
    INDEX
    Explanations

    terms related to obstruction or blockage 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
    47
    +0.12
    0.7%
    243
    +0.11
    0.6%
    389
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    56
    +0.12
    -0.01
    148
    +0.11
    0.00
    115
    +0.11
    0.02
    Negative Logits
     proud
    -1.78
     suprem
    -1.70
     publisher
    -1.66
     fair
    -1.53
    books
    -1.49
     Publ
    -1.47
     damn
    -1.47
     respectful
    -1.47
     race
    -1.45
     national
    -1.43
    POSITIVE LOGITS
    Ļª
    2.90
    ¯
    2.84
    µ
    2.83
    ĭ
    2.74
    Ĩ
    2.70
    ·¸
    2.68
    ¨
    2.67
    ·
    2.67
    Ģ
    2.66
    ®
    2.60
    Activations Density 0.353%

    No Known Activations