INDEX

Explanations

phrases indicating comparison or evaluation, focusing on the outcome or result

oai_token-act-pair · gpt-3.5-turbo

phrases that describe conditions or situations and their characteristic qualities

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

————

-0.63

MQ

-0.62

 Doing

-0.59

wheel

-0.59

SPA

-0.58

ugu

-0.58

ulk

-0.57

onday

-0.56

Whe

-0.56

ipping

-0.56

POSITIVE LOGITS

 resembles

1.14

 exceeds

1.08

 justifies

1.01

 contradicts

0.99

 mirrors

0.97

 inspires

0.95

 surpass

0.95

 undermines

0.94

 horr

0.94

 prevents

0.93

Activations Density 0.132%

phrases indicating comparison or evaluation, focusing on the outcome or result

phrases that describe conditions or situations and their characteristic qualities

No Comments

No Known Activations