INDEX

Explanations

Roman numerals and abbreviations related to official titles or institutions

oai_token-act-pair · gpt-3.5-turbo

sequences of letters and initials that form acronyms or names

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

Shape

-0.79

Materials

-0.71

poke

-0.70

TPPStreamerBot

-0.69

tips

-0.67

Pin

-0.66

mouse

-0.66

Offline

-0.66

Ã§

-0.65

tip

-0.65

POSITIVE LOGITS

.?

0.75

.,

0.67

 Seraph

0.66

#$

0.64

.:

0.61

CLASSIFIED

0.60

minster

0.59

 Angels

0.59

hyde

0.59

 Colossus

0.58

Activations Density 0.061%

Roman numerals and abbreviations related to official titles or institutions

sequences of letters and initials that form acronyms or names

No Comments

No Known Activations