© 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. 3-RES-JB
    11. 10642
    Prev
    Next
    INDEX
    Explanations

    references to specific locations or entities named "dale" with varying levels of activation

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "dale" in various contexts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 3-res-jb
    Configuration
    jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    3
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    ugal
    -0.96
    uming
    -0.77
    urers
    -0.77
    selage
    -0.77
    ntil
    -0.76
    ortunately
    -0.71
    uncture
    -0.70
    ulas
    -0.70
    ossibility
    -0.68
    htaking
    -0.67
    POSITIVE LOGITS
    dale
    1.42
    ville
    0.90
    stones
    0.88
    hill
    0.83
    mont
    0.82
    berry
    0.81
    ragon
    0.79
    phia
    0.77
    stone
    0.74
     Borough
    0.72
    Activations Density 0.010%

    No Known Activations