INDEX

Explanations

mentions of events or discoveries related to science, research, or investigations

oai_token-act-pair · gpt-3.5-turbo

occurrences of the letter 'y' 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

 Quin

-0.71

 mosqu

-0.67

handled

-0.63

 glim

-0.63

oway

-0.63

 pestic

-0.63

oter

-0.62

 Tanz

-0.61

 stocking

-0.60

asel

-0.60

POSITIVE LOGITS

utics

0.99

ISM

0.81

fulness

0.79

SHIP

0.78

bringer

0.77

anship

0.77

 flourished

0.73

 ..........

0.73

istries

0.72

 awaits

0.72

Activations Density 0.118%

mentions of events or discoveries related to science, research, or investigations

occurrences of the letter 'y' in various contexts

No Comments

No Known Activations