© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B-IT SAEs
    5. Gemma-2B-IT
    7. Residual Stream
    9. 12-RES-JB
    11. 3322
    Prev
    Next
    INDEX
    Explanations

    phrases indicating a spatial relationship between objects or individuals, such as "in front of" or "at the studio"

    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
    1678
    +0.11
    0.3%
    990
    +0.11
    0.3%
    31
    +0.10
    0.3%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    1747
    +0.11
    0.03
    990
    +0.11
    0.02
    1678
    +0.10
    0.02
    Negative Logits
     disagre
    -0.73
     apprehen
    -0.69
     unspeak
    -0.68
     indescri
    -0.67
     pymysql
    -0.67
     shenan
    -0.66
     depic
    -0.63
     Chapitre
    -0.63
     Docteur
    -0.62
    HARRY
    -0.62
    POSITIVE LOGITS
     davanti
    0.65
     nakalista
    0.63
    CloseOperation
    0.58
     frente
    0.57
     cucchia
    0.56
     autunno
    0.56
    uxxxx
    0.56
    RegressionTest
    0.56
     bevor
    0.55
    stolic
    0.54
    Activations Density 0.102%

    No Known Activations