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. 12-RES-JB
    6. 167
    Prev
    Next
    INDEX
    Explanations

    emotional expressions conveying love and relationships

    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
    +0.40
    1.7%
    2019
    +0.21
    0.9%
    736
    +0.19
    0.8%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    2019
    +0.40
    0.10
    1959
    +0.21
    0.09
    736
    +0.19
    0.07
    Negative Logits
    <bos>
    -1.82
     Kategor
    -1.71
     télécharge
    -1.64
     karton
    -1.61
     kompati
    -1.61
     Secrétaire
    -1.59
     Konkur
    -1.58
     kosme
    -1.58
     biograf
    -1.57
     stoff
    -1.54
    POSITIVE LOGITS
     pecuniary
    0.84
    rehensive
    0.79
    tiously
    0.79
     gewohnt
    0.75
     materialistic
    0.75
     snart
    0.74
    rictions
    0.74
     unfore
    0.74
    carbons
    0.73
     barbacoa
    0.73
    Activations Density 0.603%

    No Known Activations