© 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. 23267
    Prev
    Next
    INDEX
    Explanations

    words related to power, such as "power plant" or "powerful"

    oai_token-act-pair · gpt-3.5-turbo

    references to electric power or energy generation

    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
     unaff
    -0.72
    furt
    -0.70
    Els
    -0.68
    è¯
    -0.68
     skelet
    -0.67
    reddits
    -0.67
    ø
    -0.66
    ÅŁ
    -0.64
     aust
    -0.64
    azel
    -0.63
    POSITIVE LOGITS
     Power
    3.66
    Power
    2.81
     POWER
    2.46
     power
    2.40
    power
    2.39
     Powers
    2.05
     Energy
    1.47
     powers
    1.46
     Powered
    1.39
     Electricity
    1.37
    Activations Density 0.016%

    No Known Activations