INDEX

Explanations

terms related to news articles

oai_token-act-pair · gpt-3.5-turbo

references to news or political contexts

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

 Hawkins

-0.76

 Downs

-0.75

 Crow

-0.75

 Nichols

-0.75

 Oaks

-0.74

 Watkins

-0.74

 Weaver

-0.69

 Bailey

-0.69

 Michaels

-0.68

 Cary

-0.67

POSITIVE LOGITS

ANK

1.06

AFP

0.99

Islamic

0.97

Turkish

0.96

Ukraine

0.95

UTERS

0.93

 Indonesian

0.91

Turkey

0.91

ONDON

0.90

Saudi

0.89

Activations Density 0.171%

terms related to news articles

references to news or political contexts

No Comments

No Known Activations