© 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 · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    5. GPT2-Small
    7. Residual Stream
    9. 0-RES-JB
    11. 4743
    Prev
    Next
    INDEX
    Explanations

    the word "beat" with a strong emphasis

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "beat" and its variations in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 0-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    0
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     specialized
    -0.66
     conceal
    -0.63
     sensitivity
    -0.63
     lenses
    -0.62
    ocating
    -0.61
     recall
    -0.60
     CCP
    -0.60
     aut
    -0.59
     consultations
    -0.59
     inhab
    -0.59
    POSITIVE LOGITS
    beat
    4.67
    Beat
    2.33
     Beat
    1.90
     beat
    1.71
     beats
    1.42
     beaten
    1.28
     beating
    1.26
    be
    1.15
    drops
    1.02
    bart
    0.98
    Activations Density 0.008%

    No Known Activations