INDEX

Explanations

phrases related to news headlines

oai_token-act-pair · gpt-3.5-turbo

colons used in various contexts, indicating sections or topics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

ISON

-0.72

izo

-0.69

reciation

-0.69

ivable

-0.69

oser

-0.68

ozo

-0.68

authorized

-0.66

osing

-0.66

oes

-0.65

Soft

-0.64

POSITIVE LOGITS

 Photograph

0.84

Hem

0.74

 Ahead

0.74

 Michele

0.72

 Provided

0.71

 Countries

0.68

 Schwarzenegger

0.68

 Gian

0.67

 Scenes

0.67

 Bach

0.67

Activations Density 0.121%

phrases related to news headlines

colons used in various contexts, indicating sections or topics

No Comments

No Known Activations