INDEX

Explanations

locations or regions

oai_token-act-pair · gpt-3.5-turbo

references to geographic locations

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

 positives

-0.64

Hug

-0.63

 Patron

-0.63

-0.62

PDATE

-0.61

 omission

-0.61

erity

-0.60

-)

-0.60

 Puppet

-0.59

 Quote

-0.58

POSITIVE LOGITS

atche

0.81

area

0.81

vironments

0.75

 countryside

0.75

 âĶ

0.73

 enclave

0.72

 Province

0.71

 Fraz

0.71

ridor

0.70

 province

0.68

Activations Density 0.148%

locations or regions

references to geographic locations

No Comments

No Known Activations