INDEX

Explanations

words related to achievements or accomplishments

oai_token-act-pair · gpt-3.5-turbo

instances of the word "earned" indicating achievements or accomplishments

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

 compr

-0.75

 illum

-0.74

 symp

-0.62

 Ultr

-0.60

 Leilan

-0.59

":["

-0.58

mos

-0.57

 resize

-0.57

pci

-0.57

issue

-0.57

POSITIVE LOGITS

 earn

0.86

hardt

0.83

earned

0.83

 laure

0.81

 earning

0.75

Earn

0.75

chieve

0.74

 Citation

0.73

 badges

0.73

 earned

0.72

Activations Density 0.021%

words related to achievements or accomplishments

instances of the word "earned" indicating achievements or accomplishments

No Comments

No Known Activations