INDEX

Explanations

phrases related to legal and formal language

oai_token-act-pair · gpt-3.5-turbo

references to abstract concepts or symbols often associated with significance

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

 mosqu

-0.80

 Sapphire

-0.70

 Hitman

-0.70

 Registered

-0.68

 handshake

-0.66

 Cookie

-0.66

 certification

-0.65

nod

-0.64

 Finger

-0.64

 Droid

-0.63

POSITIVE LOGITS

Ĵ

1.06

¤

0.99

gat

0.96

irrel

0.93

Â§

0.91

¡

0.90

particularly

0.89

certain

0.88

rational

0.87

oras

0.87

Activations Density 0.385%

phrases related to legal and formal language

references to abstract concepts or symbols often associated with significance

No Comments

No Known Activations