INDEX

Explanations

words related to rings, such as jewelry, physical structures, and actions associated with ringing

oai_token-act-pair · gpt-3.5-turbo

instances of the word "ring" and "rings"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

theless

-0.83

heit

-0.75

 Meier

-0.72

 Agric

-0.71

DEP

-0.71

eval

-0.71

sci

-0.70

ourced

-0.70

 McCl

-0.69

 Unreal

-0.68

POSITIVE LOGITS

 ring

3.99

 rings

3.11

 Ring

2.49

Ring

2.36

 Rings

1.99

rings

1.74

ring

1.73

 amulet

1.70

 ringing

1.59

 rang

1.43

Activations Density 0.026%

words related to rings, such as jewelry, physical structures, and actions associated with ringing

instances of the word "ring" and "rings"

No Comments

No Known Activations