INDEX

Explanations

mentions of something being "backed" or supported

oai_token-act-pair · gpt-3.5-turbo

words related to physical actions or states involving "up" or "down."

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

beit

-0.65

 retri

-0.61

 warr

-0.61

 carriage

-0.60

scape

-0.59

PF

-0.59

 trave

-0.58

Sm

-0.58

 Siber

-0.57

 ancest

-0.55

POSITIVE LOGITS

olicy

1.20

ublic

0.93

rison

0.87

inion

0.83

odcast

0.82

dates

0.80

utics

0.79

osition

0.78

pping

0.77

onent

0.76

Activations Density 0.018%

mentions of something being "backed" or supported

words related to physical actions or states involving "up" or "down."

No Comments

No Known Activations