INDEX

Explanations

instances where items are mentioned or discussed in various contexts

oai_token-act-pair · gpt-3.5-turbo

references to various products or goods

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

orks

-0.82

doms

-0.74

osate

-0.72

sburgh

-0.67

 Telegram

-0.67

BILITY

-0.65

Gib

-0.63

wards

-0.63

 Doodle

-0.62

FER

-0.62

POSITIVE LOGITS

 items

0.91

urgical

0.90

meal

0.84

izer

0.81

ize

0.80

istics

0.80

 belongings

0.79

chest

0.77

ized

0.77

iser

0.76

Activations Density 0.020%

instances where items are mentioned or discussed in various contexts

references to various products or goods

No Comments

No Known Activations