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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 10-RES-JB
    11. 8413
    Prev
    Next
    INDEX
    Explanations

    phrases related to personal experiences and memories

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 10-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.10.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    10
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    soever
    -0.74
     denies
    -0.69
     observers
    -0.68
     deems
    -0.68
     surfaces
    -0.67
    uates
    -0.67
     endif
    -0.67
     factors
    -0.66
     punches
    -0.66
    RELATED
    -0.66
    POSITIVE LOGITS
     my
    1.44
     myself
    1.10
     MY
    0.94
     our
    0.94
     Hacker
    0.88
     Trader
    0.81
     reddit
    0.80
     Dept
    0.79
    my
    0.75
     mine
    0.75
    Activations Density 0.459%

    No Known Activations