INDEX

Explanations

terms related to an initial or starting point

oai_token-act-pair · gpt-3.5-turbo

instances of the word "initially."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

ake

-0.63

akings

-0.63

 impunity

-0.63

 haun

-0.61

ween

-0.60

achment

-0.60

bage

-0.59

 goddamn

-0.59

 ..."

-0.59

itia

-0.58

POSITIVE LOGITS

 initially

3.32

 originally

2.21

Initially

1.88

 Initially

1.88

 initial

1.82

 eventually

1.43

 previously

1.41

 ultimately

1.37

 momentarily

1.31

 briefly

1.29

Activations Density 0.018%

terms related to an initial or starting point

instances of the word "initially."

No Comments

No Known Activations