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-IT SAEs
    3. Gemma-2B-IT
    4. Residual Stream
    5. 12-RES-JB
    6. 1523
    Prev
    Next
    INDEX
    Explanations

    scientific research paper references

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    standard
    Context Size
    1,024
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    1343
    +0.17
    0.5%
    876
    +0.13
    0.4%
    453
    +0.12
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    780
    +0.17
    0.02
    924
    +0.13
    0.02
    1784
    +0.12
    0.02
    Negative Logits
     unspeak
    -1.25
     Shakspeare
    -1.18
     swarovski
    -1.09
     indescri
    -1.01
     gaily
    -1.01
     Whence
    -1.01
     Juf
    -1.00
     unwarran
    -0.98
     shewn
    -0.97
     inconce
    -0.96
    POSITIVE LOGITS
    <bos>
    0.87
    PhysRev
    0.68
     verkle
    0.63
     impon
    0.62
    巽
    0.56
     eleg
    0.54
     toller
    0.53
     Trieb
    0.53
     potes
    0.53
    journal
    0.53
    Activations Density 0.046%

    No Known Activations