INDEX

Explanations

details related to digging, burying, and uncovering information or objects

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

references to digging, uncovering, or exploring hidden or buried concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

vantage

-0.80

trak

-0.76

tesy

-0.71

uties

-0.71

hyde

-0.70

OPA

-0.68

capt

-0.68

owler

-0.67

Ľ

-0.65

HCR

-0.65

POSITIVE LOGITS

 rubble

1.29

 bedrock

1.24

 bones

1.24

 layers

1.23

 weeds

1.21

 fabric

1.19

 marrow

1.18

 folds

1.13

 skulls

1.13

 foundations

1.12

Activations Density 0.432%

details related to digging, burying, and uncovering information or objects

references to digging, uncovering, or exploring hidden or buried concepts

No Comments

No Known Activations