INDEX

Explanations

prefixes used in written text, such as "The" before specific nouns

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "The"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 reserve

-0.70

 wound

-0.68

 rank

-0.68

 luck

-0.66

 arch

-0.64

 drop

-0.63

 assigned

-0.63

 care

-0.63

 favor

-0.63

 equivalent

-0.62

POSITIVE LOGITS

The

2.30

There

1.72

ccording

1.72

THE

1.64

This

1.64

When

1.61

Both

1.58

It

1.58

While

1.56

Our

1.56

Activations Density 0.231%

prefixes used in written text, such as "The" before specific nouns

occurrences of the word "The"

No Comments

No Known Activations