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

    Neuronpedia

    APISteerSAE EvalsBlog/PodcastNEWSlackPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlog/PodcastGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Initial Gemma-2B SAEs
    3. Gemma-2B
    4. Residual Stream
    5. 12-RES-JB
    6. 7879
    Prev
    Next
    INDEX
    Explanations

    references to celebrity appearances and promotional events

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-Residual-Stream-SAEs/gemma_2b_blocks.12.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    torch.float32
    Hook Point
    blocks.12.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    HuggingFaceFW/fineweb
    Hook Point Layer
    12
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    50
    +0.34
    1.3%
    382
    +0.13
    0.5%
    1177
    +0.13
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1445
    +0.34
    0.11
    2019
    +0.13
    0.10
    752
    +0.13
    0.07
    Negative Logits
    <bos>
    -1.59
     keines
    -0.61
     keinerlei
    -0.61
    ,
    -0.61
     Zeilen
    -0.61
    (
    -0.61
     be
    -0.60
    /
    -0.60
     dieselben
    -0.59
    ↵↵
    -0.59
    POSITIVE LOGITS
     lele
    1.77
     Juf
    1.77
     Minang
    1.74
     Keny
    1.64
     seksi
    1.61
     Khart
    1.59
     bandung
    1.58
     Jambi
    1.55
     Palembang
    1.53
     kasa
    1.52
    Activations Density 0.828%

    No Known Activations