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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 2835
    Prev
    Next
    INDEX
    Explanations

    mathematical operations, especially those related to computer programming such as diffusions and mesh vertices

    oai_token-act-pair · gpt-3.5-turbo

    occurrences of the word "fore" or its variations in different 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
     Ogre
    -0.69
     aspir
    -0.68
     Brewer
    -0.63
     Atkins
    -0.63
     packed
    -0.61
     kW
    -0.61
     shock
    -0.59
     Alliance
    -0.59
     Brawl
    -0.58
     pedal
    -0.58
    POSITIVE LOGITS
    fore
    4.48
    FORE
    2.03
    Fore
    1.74
    forest
    1.62
    upon
    1.36
    ver
    1.24
    fort
    1.11
    cause
    1.07
    ctors
    1.02
    fter
    1.02
    Activations Density 0.009%

    No Known Activations