INDEX

Explanations

the word "lessons"

oai_token-act-pair · gpt-4-turbo

references to learning or teachings

oai_token-act-pair · gpt-3.5-turbo

repeated references to lessons learned

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

berman

-0.67

conn

-0.65

 flush

-0.62

 blob

-0.62

ractive

-0.61

riot

-0.60

secut

-0.60

 occupancy

-0.59

represented

-0.59

pub

-0.58

POSITIVE LOGITS

 lessons

1.25

 Learned

1.12

 Lessons

1.07

 lesson

0.92

 learnt

0.85

 Teach

0.84

chool

0.83

Lear

0.82

learn

0.80

ĸļ

0.78

Activations Density 0.011%

the word "lessons"

references to learning or teachings

repeated references to lessons learned

No Comments

No Known Activations