© 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. 486
    Prev
    Next
    INDEX
    Explanations

    verbs related to accepting or invading

    oai_token-act-pair · gpt-3.5-turbo

    concepts related to acceptance and invitation

    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
     finish
    -0.81
     nod
    -0.79
     pile
    -0.78
     smile
    -0.75
     layer
    -0.75
     bow
    -0.72
     pitch
    -0.72
     successor
    -0.72
     bound
    -0.72
     heights
    -0.71
    POSITIVE LOGITS
    character
    1.94
    inv
    1.86
    accept
    1.83
    program
    1.83
    engine
    1.81
    offic
    1.80
    valid
    1.78
    event
    1.75
    follow
    1.74
    introdu
    1.74
    Activations Density 0.096%

    No Known Activations