INDEX

Explanations

news mentions related to a specific media outlet called "Global Post"

oai_token-act-pair · gpt-3.5-turbo

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

 king

-0.78

liest

-0.78

 tune

-0.77

 crawl

-0.73

 routine

-0.72

 domestically

-0.71

lessly

-0.71

quo

-0.70

 enshr

-0.69

 globally

-0.69

POSITIVE LOGITS

 Affairs

1.40

 Relations

1.25

 Resources

1.17

 Strategies

1.16

 Assessment

1.16

 Sciences

1.15

 Dynamics

1.15

 Investigations

1.15

 Solutions

1.14

 Services

1.14

Activations Density 0.120%

news mentions related to a specific media outlet called "Global Post"

No Comments

No Known Activations