INDEX

Explanations

phrases or terms related to the word "Word."

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "Word" and its variations

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

 platinum

-0.75

oglu

-0.70

 capped

-0.65

 electr

-0.62

independence

-0.61

 achieve

-0.60

 commissions

-0.60

pai

-0.60

 emer

-0.59

 cartel

-0.58

POSITIVE LOGITS

 Word

3.87

Word

3.14

 word

1.97

word

1.91

 Words

1.67

words

1.32

Words

1.21

wordpress

1.09

 PowerPoint

1.04

 WordPress

0.99

Activations Density 0.015%

phrases or terms related to the word "Word."

references to the concept of "Word" and its variations

No Comments

No Known Activations