INDEX

Explanations

phrases related to comparisons or evaluations based on quality

oai_token-act-pair · gpt-3.5-turbo

negative phrases and sentiment expressions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Reloaded

-0.84

 substitution

-0.70

 Extras

-0.67

 anthem

-0.65

lication

-0.63

 behavi

-0.63

hement

-0.63

 chant

-0.62

 Nutr

-0.62

 Alley

-0.62

POSITIVE LOGITS

adequ

1.09

expected

1.02

usual

0.97

yet

0.95

necess

0.93

average

0.93

specified

0.91

mentioned

0.89

true

0.88

recomm

0.87

Activations Density 0.075%

phrases related to comparisons or evaluations based on quality

negative phrases and sentiment expressions

No Comments

No Known Activations