INDEX

Explanations

the abbreviation "OD" at different levels of activation

oai_token-act-pair · gpt-3.5-turbo

references to specific university-related terminology or acronyms

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

Ki

-0.72

Pi

-0.71

 Welsh

-0.68

 Canter

-0.67

 Hurricanes

-0.66

 Panthers

-0.65

 Pens

-0.64

 Crest

-0.64

Wem

-0.64

 Athena

-0.64

POSITIVE LOGITS

IUM

1.13

ouble

1.03

ependent

1.02

ocument

1.00

gins

1.00

CAST

0.98

OME

0.96

irect

0.95

OD

0.94

sense

0.93

Activations Density 0.009%

the abbreviation "OD" at different levels of activation

references to specific university-related terminology or acronyms

No Comments

No Known Activations