INDEX

Explanations

words ending in "-ible"

oai_token-act-pair · gpt-3.5-turbo

variations of the word "ible" or "ibles," indicating a focus on descriptors related to capability or permission

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

ster

-0.79

ugal

-0.77

ËĪ

-0.76

auri

-0.75

sters

-0.74

erva

-0.70

ority

-0.70

orius

-0.70

wu

-0.70

alyst

-0.68

POSITIVE LOGITS

ible

0.85

ibility

0.77

IBLE

0.77

vous

0.75

ibles

0.75

ibly

0.72

Mae

0.67

verning

0.66

 cords

0.65

ONSORED

0.65

Activations Density 0.021%

words ending in "-ible"

variations of the word "ible" or "ibles," indicating a focus on descriptors related to capability or permission

No Comments

No Known Activations

words ending in "-ible"

variations of the word "ible" or "ibles," indicating a focus on descriptors related to capability or permission

No Comments

No Known Activations