INDEX

Explanations

terms related to data retention and attrition

oai_token-act-pair · gpt-3.5-turbo

terms associated with data retention and management

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

enegger

-0.71

 Billion

-0.66

Ô

-0.65

ologue

-0.64

externalActionCode

-0.64

kef

-0.63

Electric

-0.61

Jenn

-0.61

Eth

-0.61

gling

-0.60

POSITIVE LOGITS

 retention

1.16

 lapse

0.91

omore

0.79

 vigilance

0.76

iple

0.73

 aversion

0.73

 readiness

0.72

luster

0.71

umatic

0.70

enh

0.70

Activations Density 0.011%

terms related to data retention and attrition

terms associated with data retention and management

No Comments

No Known Activations