INDEX

Explanations

past actions or habits

oai_token-act-pair · gpt-3.5-turbo

phrases indicating past experiences or actions using the phrase "used to."

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

 ensued

-0.68

worthiness

-0.67

ouf

-0.65

abl

-0.64

 deserves

-0.63

ographies

-0.62

lip

-0.61

worthy

-0.61

icious

-0.60

icipated

-0.60

POSITIVE LOGITS

igslist

0.79

 rent

0.78

 shrink

0.75

 hear

0.75

 bombard

0.75

 bury

0.74

 find

0.73

 advertise

0.72

 form

0.72

peg

0.72

Activations Density 0.057%

past actions or habits

phrases indicating past experiences or actions using the phrase "used to."

No Comments

No Known Activations