© 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. 11-RES-JB
    11. 2446
    Prev
    Next
    INDEX
    Explanations

    phrases indicating an authority figure giving commands or directives

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "ordered" related to directives or commands

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 11-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    11
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    nown
    -0.94
    pmwiki
    -0.92
    fact
    -0.72
    uded
    -0.72
    angular
    -0.68
    rift
    -0.67
     sidx
    -0.67
    fortunately
    -0.67
    alth
    -0.66
    leading
    -0.66
    POSITIVE LOGITS
     subordinates
    0.93
     immediate
    0.90
     demolition
    0.87
     removal
    0.81
     destruction
    0.78
     deletion
    0.78
     evacuation
    0.78
     expulsion
    0.77
     him
    0.77
     extermin
    0.76
    Activations Density 0.069%

    No Known Activations