INDEX

Explanations

phrases related to instructions or directives

oai_token-act-pair · gpt-3.5-turbo

phrases indicating informative content or announcements

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

sed

-0.62

lication

-0.62

olean

-0.61

onse

-0.60

 bragging

-0.59

itiz

-0.58

minist

-0.58

 elim

-0.57

"""

-0.57

 outl

-0.57

POSITIVE LOGITS

 HuffPost

0.95

Featured

0.95

iframe

0.90

 Coverage

0.89

 slideshow

0.86

itbart

0.85

Source

0.84

Alert

0.83

Correction

0.82

 Update

0.82

Activations Density 0.324%

phrases related to instructions or directives

phrases indicating informative content or announcements

No Comments

No Known Activations