INDEX

Explanations

terms related to size and scale

oai_token-act-pair · gpt-3.5-turbo

phrases related to directions or guidance

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

unes

-0.67

Decre

-0.67

 Refugees

-0.67

berries

-0.66

 Clothing

-0.65

ãĥĨ

-0.64

prev

-0.64

BO

-0.64

 Lowry

-0.64

neutral

-0.63

POSITIVE LOGITS

 larger

1.04

 sprawling

0.99

 elaborate

0.97

 bigger

0.97

 massive

0.96

 broader

0.95

 complex

0.94

 systemic

0.94

 sophisticated

0.92

 deeper

0.92

Activations Density 0.371%

terms related to size and scale

phrases related to directions or guidance

No Comments

No Known Activations