© 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. 4-RES-JB
    11. 13194
    Prev
    Next
    INDEX
    Explanations

    expressions related to granting permission or authority

    oai_token-act-pair · gpt-3.5-turbo

    terms related to power, authority, and autonomy

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

    No Comments

    Negative Logits
     LCD
    -0.65
     bicycl
    -0.64
    MT
    -0.64
     enthusi
    -0.64
     Peb
    -0.63
    Interstitial
    -0.62
    unct
    -0.61
    ronics
    -0.61
    LINE
    -0.61
    raph
    -0.60
    POSITIVE LOGITS
     granted
    0.90
     privileges
    0.89
     allotted
    0.87
     immunity
    0.82
     rights
    0.82
    hold
    0.81
    license
    0.80
     permissions
    0.80
     guarantee
    0.76
    aterasu
    0.76
    Activations Density 0.394%

    No Known Activations