INDEX

Explanations

instances where there is an increase or growth in something

oai_token-act-pair · gpt-3.5-turbo

instances of the word "increase" and its variations in the context of growth or escalation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

aries

-0.75

thing

-0.74

mates

-0.74

rome

-0.67

ician

-0.66

stones

-0.66

gets

-0.66

halla

-0.65

atom

-0.65

 Monstrous

-0.63

POSITIVE LOGITS

 exponentially

1.05

 visibility

1.03

 awareness

1.01

 likelihood

0.95

 efficiency

0.91

 amounts

0.87

 transparency

0.87

 effic

0.87

 sensitivity

0.86

 throughput

0.83

Activations Density 0.071%

instances where there is an increase or growth in something

instances of the word "increase" and its variations in the context of growth or escalation

No Comments

No Known Activations