INDEX

Explanations

phrases or statements indicating a continuous or ongoing action or event

oai_token-act-pair · gpt-3.5-turbo

the word "continues" and its variations, indicating ongoing actions or states

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

Ã©s

-0.82

iah

-0.79

ool

-0.78

ographical

-0.71

ustomed

-0.71

angs

-0.71

atorial

-0.70

anged

-0.69

ooters

-0.69

arta

-0.69

POSITIVE LOGITS

 unab

0.99

ap

0.81

 uninterrupted

0.78

 onward

0.77

 unanswered

0.74

 unchanged

0.72

 unsolved

0.68

 tremend

0.68

 unresolved

0.67

 indefinitely

0.66

Activations Density 0.040%

phrases or statements indicating a continuous or ongoing action or event

the word "continues" and its variations, indicating ongoing actions or states

No Comments

No Known Activations