INDEX

Explanations

references to specific episodes or titles of a show or game series

oai_token-act-pair · gpt-3.5-turbo

references to specific anime and video game titles

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 eleph

-0.78

bush

-0.72

miss

-0.71

 spons

-0.70

tackle

-0.70

hower

-0.69

rup

-0.67

conserv

-0.67

onder

-0.65

 Krug

-0.65

POSITIVE LOGITS

acters

0.90

 Characters

0.87

 encyclopedia

0.81

 Chronicles

0.80

 sequels

0.79

 Generations

0.79

 videog

0.78

 trilogy

0.78

 sequel

0.78

 manga

0.77

Activations Density 0.065%

references to specific episodes or titles of a show or game series

references to specific anime and video game titles

No Comments

No Known Activations