Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Joseph Bloom · Open Source Sparse Autoencoders for all Residual Stream Layers of GPT2-Small
    3. GPT2-Small
    4. Residual Stream
    5. 0-RES-JB
    6. 5800
    Prev
    Next
    INDEX
    Explanations

    words related to correspondence or reporting

    oai_token-act-pair · gpt-3.5-turbo

    references to correspondents and their roles in various contexts

    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
    sterdam
    -0.79
    ixel
    -0.77
    glass
    -0.71
    asonic
    -0.71
    azon
    -0.70
    teenth
    -0.68
    ths
    -0.68
    anco
    -0.65
    ragon
    -0.64
    anan
    -0.64
    POSITIVE LOGITS
     Correspond
    3.52
     Situation
    2.08
    respond
    1.12
     Surviv
    0.88
     Cous
    0.84
     condol
    0.81
     Princ
    0.80
     Reported
    0.78
    pected
    0.77
     Deliver
    0.76
    Activations Density 0.022%

    No Known Activations