INDEX

Explanations

words related to social impact or interactions within communities

oai_token-act-pair · gpt-3.5-turbo

references to impacts related to cannabis usage and mental health

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Lauder

-0.67

IU

-0.62

 DeVos

-0.61

 Sack

-0.61

 Kavanaugh

-0.59

 Applicant

-0.59

 Akin

-0.57

akia

-0.57

 Zika

-0.55

Asset

-0.55

POSITIVE LOGITS

 everyday

0.65

 unsuspecting

0.63

aneously

0.61

 geographically

0.60

 terrestrial

0.60

 toast

0.57

 mainstream

0.56

 persecuted

0.55

[|

0.55

inates

0.55

Activations Density 1.441%

words related to social impact or interactions within communities

references to impacts related to cannabis usage and mental health

No Comments

No Known Activations