© 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. 1710
    Prev
    Next
    INDEX
    Explanations

    words related to forwarding or transmission of information

    oai_token-act-pair · gpt-3.5-turbo

    the word "forwarded" and its variations, indicating a focus on email or information forwarding

    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
    teenth
    -0.86
    gins
    -0.82
    ENE
    -0.77
    aila
    -0.76
    abase
    -0.75
    teen
    -0.75
    enez
    -0.75
    âĶĢ
    -0.75
     Pett
    -0.74
    BuyableInstoreAndOnline
    -0.70
    POSITIVE LOGITS
    forward
    1.12
     forwarding
    0.96
     Forward
    0.80
     forwarded
    0.80
    Forward
    0.77
    xual
    0.70
    intendent
    0.69
    folios
    0.66
     Generations
    0.66
    Iterator
    0.64
    Activations Density 0.018%

    No Known Activations