INDEX

Explanations

terms related to size or magnitude

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "scale" 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 @ 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

esson

-0.89

odge

-0.80

ĪĴ

-0.77

 Leilan

-0.73

aan

-0.72

ton

-0.72

ournal

-0.69

tein

-0.69

phis

-0.68

nell

-0.68

POSITIVE LOGITS

 scale

0.93

enter

0.75

scale

0.72

 fabrication

0.71

Factor

0.68

 scales

0.68

enance

0.67

 proportions

0.67

 cooker

0.66

 invari

0.65

Activations Density 0.016%

terms related to size or magnitude

references to the concept of "scale" in various contexts

No Comments

No Known Activations