INDEX

Explanations

words related to medical imaging technology, specifically CT scans

oai_token-act-pair · gpt-3.5-turbo

mentions of Connecticut

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

wagen

-0.88

leased

-0.79

volent

-0.79

ĺħ

-0.77

conservancy

-0.74

 myster

-0.72

hot

-0.71

Pwr

-0.69

doms

-0.67

mentioned

-0.67

POSITIVE LOGITS

ARDIS

0.90

RL

0.89

ICLE

0.85

urtle

0.80

arre

0.80

arette

0.80

INO

0.77

ivity

0.76

ocular

0.75

CT

0.75

Activations Density 0.013%

words related to medical imaging technology, specifically CT scans

mentions of Connecticut

No Comments

No Known Activations