INDEX

Explanations

terms related to active engagement, support, and involvement

oai_token-act-pair · gpt-3.5-turbo

the word "actively" and its variations in contexts related to engagement or involvement

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

 Parables

-0.92

elf

-0.81

esan

-0.75

 Chaser

-0.68

Fine

-0.66

wich

-0.66

 Arri

-0.65

 Tycoon

-0.64

 Horizons

-0.64

nian

-0.64

POSITIVE LOGITS

 engaged

1.10

uated

0.86

uating

0.84

 encouraged

0.83

 participated

0.83

 actively

0.80

 solic

0.79

 participates

0.79

 discouraged

0.79

 recruited

0.78

Activations Density 0.012%

terms related to active engagement, support, and involvement

the word "actively" and its variations in contexts related to engagement or involvement

No Comments

No Known Activations