INDEX

Explanations

veins and arteries in medical contexts

oai_token-act-pair · gpt-3.5-turbo

terms related to veins and arteries

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

ordan

-0.67

-+-+-+-+

-0.64

eur

-0.64

*/(

-0.64

MAT

-0.63

rooms

-0.63

Gy

-0.61

ommod

-0.61

%]

-0.61

Ħ¢

-0.60

POSITIVE LOGITS

 vein

1.29

 veins

0.96

phrine

0.91

marks

0.82

ectomy

0.82

itial

0.80

osure

0.75

uous

0.73

clave

0.72

cles

0.71

Activations Density 0.012%

veins and arteries in medical contexts

terms related to veins and arteries

No Comments

No Known Activations