© 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. 10-RES-JB
    11. 21646
    Prev
    Next
    INDEX
    Explanations

    phrases related to actions affecting people or group of people, like being kept, made, turned, written off, brought down, and set up

    oai_token-act-pair · gpt-3.5-turbo

    actions related to being constrained or at a loss of autonomy

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    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
     Stability
    -0.69
    rium
    -0.69
     EF
    -0.65
     genesis
    -0.65
     Walls
    -0.63
     Forge
    -0.62
     Pillar
    -0.62
     Edwards
    -0.61
     ECO
    -0.60
     Rivers
    -0.60
    POSITIVE LOGITS
     aback
    1.12
     unfairly
    0.88
     alive
    0.85
     captive
    0.84
    quished
    0.84
    ependent
    0.83
    aucas
    0.83
     hostage
    0.82
    otyp
    0.82
    gyn
    0.81
    Activations Density 0.242%

    No Known Activations