INDEX

Explanations

words related to medical terms, particularly medical conditions and procedures

oai_token-act-pair · gpt-3.5-turbo

words related to medical or scientific terminology

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

staking

-0.67

fully

-0.65

-0.64

fulness

-0.63

 count

-0.61

 reality

-0.61

anwhile

-0.60

PDATE

-0.59

eve

-0.58

 condensed

-0.58

POSITIVE LOGITS

atform

1.24

asso

1.10

asts

1.05

asms

1.02

asma

1.00

ifting

0.97

utonium

0.95

opl

0.93

ocate

0.93

asm

0.91

Activations Density 0.010%

words related to medical terms, particularly medical conditions and procedures

words related to medical or scientific terminology

No Comments

No Known Activations