INDEX

Explanations

phrases related to prevention and control of various issues, particularly diseases and crimes

oai_token-act-pair · gpt-3.5-turbo

terms related to prevention, particularly in health and safety contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

urn

-0.75

 likeness

-0.69

 churn

-0.59

 towering

-0.59

 upload

-0.59

fell

-0.58

orrow

-0.58

 borrow

-0.58

 owes

-0.57

plex

-0.57

POSITIVE LOGITS

 Prevention

3.84

 prevention

2.94

 preventive

1.77

 Prevent

1.58

 mitigation

1.25

 Protection

1.18

 Detection

1.16

 Reduction

1.15

 avoidance

1.06

 Epidem

1.04

Activations Density 0.019%

phrases related to prevention and control of various issues, particularly diseases and crimes

terms related to prevention, particularly in health and safety contexts

No Comments

No Known Activations