INDEX

Explanations

references to different departments or agencies of health in various locations

oai_token-act-pair · gpt-3.5-turbo

references to health-related institutions and organizations

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

ounter

-0.75

ptions

-0.74

plets

-0.72

oxide

-0.71

cair

-0.69

 Gibbs

-0.67

 slee

-0.67

plet

-0.67

nces

-0.66

ãĥ£

-0.66

POSITIVE LOGITS

 Care

1.15

care

1.10

Care

1.09

osate

0.93

aceutical

0.87

 Insurance

0.86

Health

0.86

iest

0.86

 care

0.85

 Sciences

0.79

Activations Density 0.026%

references to different departments or agencies of health in various locations

references to health-related institutions and organizations

No Comments

No Known Activations