INDEX

Explanations

words related to the action of using something or applying a method

oai_token-act-pair · gpt-3.5-turbo

instances of the word "use."

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

 presided

-0.70

 amidst

-0.69

 Born

-0.65

jri

-0.63

appointed

-0.61

quo

-0.60

hips

-0.60

bled

-0.59

pher

-0.58

iah

-0.58

POSITIVE LOGITS

use

3.27

Use

2.37

USE

2.25

Use

2.17

use

2.10

 uses

2.09

 usage

2.06

 reuse

1.94

 misuse

1.89

 utilize

1.89

Activations Density 0.053%

words related to the action of using something or applying a method

instances of the word "use."

No Comments

No Known Activations