INDEX

Explanations

words related to completion or execution of tasks or actions

oai_token-act-pair · gpt-3.5-turbo

phrases indicating actions or processes that have been completed or executed

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Flavoring

-0.75

 Founding

-0.72

ories

-0.64

gdala

-0.62

usha

-0.61

 Flavor

-0.61

orum

-0.59

ses

-0.59

assi

-0.59

 Ratings

-0.58

POSITIVE LOGITS

 purposefully

1.07

 anonymously

1.07

 unconsciously

1.04

 deliberately

1.04

 democratically

1.03

 consciously

1.02

 electronically

1.01

 solely

1.00

 expressly

0.98

 concurrently

0.98

Activations Density 0.227%

words related to completion or execution of tasks or actions

phrases indicating actions or processes that have been completed or executed

No Comments

No Known Activations