INDEX

Explanations

words related to leaks or information disclosure

oai_token-act-pair · gpt-3.5-turbo

variations of the word "leak" and its related forms

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

 Frie

-0.76

osh

-0.76

allic

-0.76

 DRAG

-0.71

ript

-0.70

ALLY

-0.70

iage

-0.68

psc

-0.66

ãĥ

-0.65

Ī

-0.62

POSITIVE LOGITS

nesses

0.92

doms

0.91

ware

0.81

hawks

0.78

pedia

0.78

 peek

0.78

Torrent

0.73

 Insider

0.72

hole

0.71

holes

0.70

Activations Density 0.054%

words related to leaks or information disclosure

variations of the word "leak" and its related forms

No Comments

No Known Activations