INDEX

Explanations

acronyms or abbreviations followed by numbers, specifically focusing on the term "LP."

oai_token-act-pair · gpt-3.5-turbo

references to albums and their classification

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

abad

-0.94

bles

-0.77

saf

-0.73

bear

-0.71

ãĤ±

-0.71

bang

-0.70

flies

-0.70

bled

-0.67

otin

-0.67

ding

-0.67

POSITIVE LOGITS

LP

1.14

rint

1.07

ropri

0.94

FS

0.91

oyd

0.88

VT

0.88

JV

0.87

FP

0.83

olicy

0.82

arser

0.82

Activations Density 0.006%

acronyms or abbreviations followed by numbers, specifically focusing on the term "LP."

references to albums and their classification

No Comments

No Known Activations