INDEX

Explanations

words related to engineering disciplines

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Eng" in various contexts

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

DragonMagazine

-0.93

 Leilan

-0.80

pmwiki

-0.72

TPPStreamerBot

-0.72

 millenn

-0.67

IFIED

-0.66

 Tale

-0.66

etheless

-0.64

 Discrimination

-0.64

istically

-0.63

POSITIVE LOGITS

lishes

1.24

raved

1.10

ravings

1.08

rish

1.00

agement

0.94

iment

0.85

raham

0.85

inse

0.84

ined

0.83

iments

0.83

Activations Density 0.009%

words related to engineering disciplines

instances of the word "Eng" in various contexts

No Comments

No Known Activations