INDEX

Explanations

medical institutions or related terms

oai_token-act-pair · gpt-3.5-turbo

references to medical institutions or entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

xual

-0.97

effective

-0.77

arget

-0.73

upon

-0.73

gger

-0.73

invoke

-0.72

ailable

-0.68

 spirited

-0.67

spir

-0.67

ounters

-0.67

POSITIVE LOGITS

 Medical

1.09

Medical

0.91

 medical

0.89

 Care

0.87

 practition

0.87

 Sciences

0.84

 Surgery

0.83

 Advisory

0.81

 Insurance

0.81

medical

0.80

Activations Density 0.006%

medical institutions or related terms

references to medical institutions or entities

No Comments

No Known Activations