INDEX

Explanations

phrases related to long-term or extensive activities or events

oai_token-act-pair · gpt-3.5-turbo

references to various types of journeys and operations

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 positives

-0.72

amples

-0.72

 exceptions

-0.70

 slightest

-0.69

addons

-0.68

 chances

-0.67

iaries

-0.66

nor

-0.66

 stickers

-0.66

 drawbacks

-0.64

POSITIVE LOGITS

 spanning

1.23

 consisting

1.18

 dubbed

1.10

 culminating

1.09

 involving

1.09

 aimed

1.07

 centered

1.05

 encomp

1.03

 spearheaded

1.03

 reminiscent

1.02

Activations Density 0.447%

phrases related to long-term or extensive activities or events

references to various types of journeys and operations

No Comments

No Known Activations