INDEX

Explanations

phrases including the word "Hi"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Hi" in various forms and contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Awakens

-0.85

å£«

-0.77

 rall

-0.74

*/(

-0.73

 destro

-0.73

 Gleaming

-0.71

icate

-0.70

edIn

-0.67

å¥³

-0.66

SHIP

-0.66

POSITIVE LOGITS

earch

0.87

Fi

0.83

pping

0.80

roy

0.78

pped

0.75

ya

0.74

Bs

0.73

dden

0.70

annis

0.70

agar

0.69

Activations Density 0.012%

phrases including the word "Hi"

instances of the word "Hi" in various forms and contexts

No Comments

No Known Activations