INDEX

Explanations

references to division or separation

oai_token-act-pair · gpt-3.5-turbo

instances of the word "split" 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

Adv

-0.80

jamin

-0.80

trak

-0.78

ugu

-0.70

FORE

-0.69

TN

-0.67

externalActionCode

-0.67

psc

-0.67

esa

-0.65

arde

-0.65

POSITIVE LOGITS

 split

0.98

SPL

0.92

Split

0.90

 splits

0.89

 hairs

0.85

ters

0.83

split

0.79

 Split

0.77

 evenly

0.75

olitan

0.74

Activations Density 0.012%

references to division or separation

instances of the word "split" in various contexts

No Comments

No Known Activations