INDEX

Explanations

numbers with a specific mathematical operation applied "+", as well as collections or groups indicated by the term "plus"

oai_token-act-pair · gpt-3.5-turbo

phrases indicating quantities or numerical values

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

rog

-0.77

 Bates

-0.71

ked

-0.66

atography

-0.65

rium

-0.65

bird

-0.65

 Hamm

-0.64

Lar

-0.64

YN

-0.64

rius

-0.63

POSITIVE LOGITS

ottest

0.77

 percentile

0.69

 IMAGES

0.69

verage

0.68

 intersections

0.67

 overload

0.67

mph

0.66

 servings

0.65

 ILCS

0.65

 consecutive

0.65

Activations Density 0.033%

numbers with a specific mathematical operation applied "+", as well as collections or groups indicated by the term "plus"

phrases indicating quantities or numerical values

No Comments

No Known Activations