INDEX

Explanations

words related to scaling, resizing, or adjusting something

oai_token-act-pair · gpt-3.5-turbo

terms related to scaling and scalability in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.hook_resid_pre

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

vous

-0.77

WP

-0.69

Mub

-0.68

naire

-0.67

roit

-0.67

 Marcos

-0.65

 Lauder

-0.65

 Guilty

-0.63

 Thur

-0.62

cond

-0.62

POSITIVE LOGITS

 scaled

0.94

 scales

0.82

itized

0.81

 scale

0.80

 heights

0.76

 scaling

0.76

Factor

0.75

 invari

0.74

 proportions

0.73

 fins

0.73

Activations Density 0.010%

words related to scaling, resizing, or adjusting something

terms related to scaling and scalability in various contexts

No Comments

No Known Activations