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

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Initial Gemma-2B SAEs
    3. Gemma-2B
    4. Residual Stream
    5. 6-RES-JB
    6. 6725
    Prev
    Next
    INDEX
    Explanations

    mentions of opera houses, operas, composers, librettists, performances, and awards related to opera

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.6.hook_resid_post_16384_anthropic_fast_lr
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    torch.float32
    Hook Point
    blocks.6.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    HuggingFaceFW/fineweb
    Hook Point Layer
    6
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    1296
    +0.16
    0.7%
    900
    +0.13
    0.6%
    31
    +0.13
    0.6%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    783
    +0.16
    0.04
    1798
    +0.13
    0.04
    990
    +0.13
    0.03
    Negative Logits
    <bos>
    -1.85
     thu
    -0.71
    outs
    -0.67
     مرئيه
    -0.66
     tụ
    -0.66
    Thành
    -0.66
    ValueGeneration
    -0.66
     частности
    -0.65
    υ
    -0.64
    addComponent
    -0.63
    POSITIVE LOGITS
     increa
    2.01
     impra
    1.98
     scrat
    1.85
     maneu
    1.84
     reluct
    1.80
     thut
    1.78
     suscep
    1.76
     snoopy
    1.75
     indestru
    1.74
     fta
    1.73
    Activations Density 0.387%

    No Known Activations