INDEX

Explanations

words related to emotions or deep feelings, specifically related to the heart

oai_token-act-pair · gpt-3.5-turbo

references to "hearts" in emotional contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

Attorney

-0.69

 Sweeney

-0.64

 aggress

-0.63

 Discrimination

-0.62

dq

-0.62

 traumatic

-0.61

pmwiki

-0.60

 outpatient

-0.60

 antidepressants

-0.60

ress

-0.58

POSITIVE LOGITS

 hearts

1.17

chool

1.02

mith

0.94

strings

0.85

pring

0.85

pots

0.84

core

0.82

ight

0.81

borough

0.79

endor

0.78

Activations Density 0.006%

words related to emotions or deep feelings, specifically related to the heart

references to "hearts" in emotional contexts

No Comments

No Known Activations