INDEX

Explanations

tweets discussing various current events and news stories

oai_token-act-pair · gpt-3.5-turbo

phrases or elements related to online content, particularly images or posts identified by "pic."

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

 induct

-0.78

 awakening

-0.76

 involuntary

-0.75

 acquisitions

-0.70

 audits

-0.70

 depreciation

-0.67

 judgments

-0.67

 multiplying

-0.66

 defe

-0.66

 bonding

-0.66

POSITIVE LOGITS

twitter

1.57

facebook

1.10

imgur

1.01

blogspot

0.96

cdn

0.96

gallery

0.94

google

0.94

github

0.92

apple

0.91

nz

0.90

Activations Density 0.035%

tweets discussing various current events and news stories

phrases or elements related to online content, particularly images or posts identified by "pic."

No Comments

No Known Activations