INDEX

Explanations

specific numbers and technical terms related to computer processors

oai_token-act-pair · gpt-3.5-turbo

references to numerical data, particularly related to years and products

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

gets

-0.75

minded

-0.72

chell

-0.70

 fuzzy

-0.64

ivities

-0.63

Vik

-0.63

ezvous

-0.62

heast

-0.62

ainment

-0.62

 yield

-0.61

POSITIVE LOGITS

ULT

0.95

0.83

pmwiki

0.82

rpm

0.81

0.80

0.79

Activations Density 0.053%

specific numbers and technical terms related to computer processors

references to numerical data, particularly related to years and products

No Comments

No Known Activations