INDEX

Explanations

terms related to hidden or unnoticed things

oai_token-act-pair · gpt-3.5-turbo

terms associated with recognition, exploitation, and difficulties related to visibility or acknowledgment

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 marching

-0.71

 spirits

-0.71

 tone

-0.71

 decor

-0.69

 redes

-0.69

 moons

-0.67

Reviewer

-0.67

 dressing

-0.67

 prisoner

-0.64

 culture

-0.64

POSITIVE LOGITS

ibly

1.39

ivably

1.34

itive

1.26

itional

1.24

ably

1.21

itionally

1.20

itable

1.17

itability

1.15

ibles

1.14

ceivable

1.11

Activations Density 0.058%

terms related to hidden or unnoticed things

terms associated with recognition, exploitation, and difficulties related to visibility or acknowledgment

No Comments

No Known Activations