INDEX

Explanations

instances of the word "how" along with actions or decision-making processes that follow afterwards

oai_token-act-pair · gpt-3.5-turbo

questions and phrases related to methods or approaches to achieving specific goals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

krit

-0.76

miah

-0.72

mingham

-0.67

idon

-0.63

 Originally

-0.61

odium

-0.61

 spite

-0.60

ãĤ©

-0.60

rored

-0.60

obyl

-0.59

POSITIVE LOGITS

to

1.20

 best

1.08

to

0.84

To

0.81

 should

0.76

soever

0.75

 thereto

0.74

 constitutes

0.73

To

0.68

TO

0.67

Activations Density 0.142%

instances of the word "how" along with actions or decision-making processes that follow afterwards

questions and phrases related to methods or approaches to achieving specific goals

No Comments

No Known Activations