INDEX

Explanations

interactions or components that are responsive and user-engaging, often in the form of interactive features or tools

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "interactive" in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ioxide

-0.93

ighed

-0.76

oil

-0.75

iffs

-0.73

aimon

-0.72

union

-0.71

arah

-0.70

hood

-0.68

arest

-0.68

yards

-0.68

POSITIVE LOGITS

 interactive

0.89

 slideshow

0.81

 multimedia

0.75

 storytelling

0.73

 theatre

0.70

 touchscreen

0.67

tee

0.66

 theater

0.66

 barrier

0.65

 underwater

0.65

Activations Density 0.014%

interactions or components that are responsive and user-engaging, often in the form of interactive features or tools

occurrences of the term "interactive" in various contexts

No Comments

No Known Activations