© 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. 11248
    Prev
    Next
    INDEX
    Explanations

    phrases related to breaking through obstacles or barriers

    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
    115
    +0.15
    0.8%
    164
    +0.13
    0.7%
    317
    +0.11
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    115
    +0.15
    0.03
    164
    +0.13
    0.03
    62
    +0.11
    0.02
    Negative Logits
    ĸ
    -2.09
    Ĥ
    -1.96
    ¢
    -1.91
    ī
    -1.84
    ¦
    -1.71
    ²
    -1.70
    ¾
    -1.66
     facto
    -1.66
    ©
    -1.65
    ¯
    -1.64
    POSITIVE LOGITS
    neck
    2.15
    point
    2.10
    points
    1.95
    tons
    1.89
    hearted
    1.83
    ups
    1.75
    away
    1.70
    gers
    1.68
    iday
    1.64
    ger
    1.64
    Activations Density 0.252%

    No Known Activations