INDEX

Explanations

proper nouns or names of entities such as countries, organizations, or specific laws

oai_token-act-pair · gpt-3.5-turbo

highly impactful statements or events in news-related contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

 undecided

-0.80

 blocker

-0.74

 democrat

-0.73

 persuasion

-0.71

guy

-0.70

 mustache

-0.70

 waterfall

-0.69

 deflation

-0.69

 grapp

-0.69

 sails

-0.68

POSITIVE LOGITS

Interested

1.57

According

1.54

The

1.46

Writing

1.41

Sources

1.38

Speaking

1.34

Instead

1.32

Following

1.32

In

1.30

While

1.30

Activations Density 0.265%

proper nouns or names of entities such as countries, organizations, or specific laws

highly impactful statements or events in news-related contexts

No Comments

No Known Activations