INDEX

Explanations

words or phrases prefixed with "pre-"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the term "pre-" as a prefix or related form

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

Nig

-0.71

 Holding

-0.70

 Cohn

-0.68

 shards

-0.68

 streams

-0.67

 DRAG

-0.67

 SHARES

-0.67

 fragments

-0.63

 Stranger

-0.63

 merits

-0.63

POSITIVE LOGITS

existing

1.65

empt

1.46

requisite

1.38

requisites

1.36

condition

1.26

trial

1.26

release

1.25

selected

1.25

existent

1.21

installed

1.21

Activations Density 0.024%

words or phrases prefixed with "pre-"

occurrences of the term "pre-" as a prefix or related form

No Comments

No Known Activations