INDEX

Explanations

mentions of something being newly released or launched

oai_token-act-pair · gpt-3.5-turbo

instances of the word "newly" associated with new releases, launches, or creations

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

lua

-0.76

alos

-0.75

ilogy

-0.72

adj

-0.70

AIDS

-0.70

 Clicker

-0.69

OTO

-0.69

imar

-0.67

olson

-0.67

ainer

-0.67

POSITIVE LOGITS

 ones

0.73

 versions

0.72

 masses

0.66

 duck

0.63

 African

0.62

 lengths

0.62

 joint

0.62

 version

0.61

 OECD

0.61

 portions

0.61

Activations Density 0.129%

mentions of something being newly released or launched

instances of the word "newly" associated with new releases, launches, or creations

No Comments

No Known Activations