INDEX

Explanations

words related to shared collective experiences

oai_token-act-pair · gpt-3.5-turbo

the phrase "we all."

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

ahime

-0.69

aminer

-0.65

âĵĺ

-0.64

plete

-0.63

luster

-0.62

srf

-0.61

 rarity

-0.61

 Lancet

-0.61

isoft

-0.60

ciation

-0.60

POSITIVE LOGITS

ocating

1.15

ocated

0.87

 smiles

0.85

owed

0.81

iances

0.80

edged

0.79

 kinds

0.78

 sorts

0.76

ende

0.75

owing

0.75

Activations Density 0.053%

words related to shared collective experiences

the phrase "we all."

No Comments

No Known Activations