INDEX

Explanations

titles of movies, books, and TV shows with descriptive adjectives

oai_token-act-pair · gpt-3.5-turbo

titles of notable films and series

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 opponents

-0.72

TN

-0.72

itionally

-0.68

 positioning

-0.66

 waivers

-0.66

 compliance

-0.66

 contention

-0.64

 credibility

-0.63

 livelihood

-0.62

ruary

-0.62

POSITIVE LOGITS

 Awakens

1.14

 Trilogy

1.04

 Angels

0.93

 Bride

0.89

 Below

0.87

 Avenger

0.86

 Swordsman

0.85

 Incident

0.85

ateful

0.84

 Album

0.83

Activations Density 0.280%

titles of movies, books, and TV shows with descriptive adjectives

titles of notable films and series

No Comments

No Known Activations