INDEX

Explanations

links or calls to action indicated by the word "here"

oai_token-act-pair · gpt-3.5-turbo

repeated phrases instructing readers to click links for additional content or information

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

itialized

-0.81

upuncture

-0.68

arthed

-0.67

inventoryQuantity

-0.64

catentry

-0.62

iour

-0.62

inguished

-0.61

TPPStreamerBot

-0.60

ONSORED

-0.60

ethy

-0.60

POSITIVE LOGITS

tics

1.06

abouts

0.92

tical

0.86

 LINK

0.75

 Submit

0.73

 link

0.72

 âĢº

0.72

tic

0.71

 links

0.69

[+]

0.68

Activations Density 0.013%

links or calls to action indicated by the word "here"

repeated phrases instructing readers to click links for additional content or information

No Comments

No Known Activations