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. 15530
    Prev
    Next
    INDEX
    Explanations

    phrases related to breaking or dismantling something

    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
    61
    +0.19
    0.7%
    1805
    +0.12
    0.4%
    1026
    +0.11
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    61
    +0.19
    0.06
    1805
    +0.12
    0.04
    976
    +0.11
    0.04
    Negative Logits
    ValueStyle
    -0.50
    Pfalz
    -0.44
    VolleyError
    -0.44
    racle
    -0.44
    LabelTagHelper
    -0.43
    AndroidJUnit
    -0.42
    AllowAnonymous
    -0.41
    czywiście
    -0.41
    FlatStyle
    -0.41
    zmir
    -0.41
    POSITIVE LOGITS
     break
    1.16
     Breaking
    1.11
     breaking
    1.10
     Break
    1.08
    break
    1.08
     breaks
    1.07
     broke
    1.04
     Breaks
    1.03
     BREAK
    1.03
    Break
    1.02
    Activations Density 0.129%

    No Known Activations