INDEX

Explanations

words in a non-English language or a different character encoding

oai_token-act-pair · gpt-3.5-turbo

specific characters or symbols in a non-English language context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

ttes

-0.84

 Starr

-0.77

otle

-0.73

ellen

-0.71

 Somers

-0.69

ulhu

-0.68

eller

-0.67

 McMaster

-0.65

 Pearce

-0.65

 Gutenberg

-0.64

POSITIVE LOGITS

ÑĤ

1.17

Ðº

1.15

Ñģ

1.05

Ð½

0.95

Ð

0.93

Ñı

0.92

Ð¼

0.92

Ð»

0.90

Ð¸

0.87

ãĥª

0.85

Activations Density 0.006%

words in a non-English language or a different character encoding

specific characters or symbols in a non-English language context

No Comments

No Known Activations