INDEX

Explanations

phrases indicating a specific action or event that is part of a broader process or investigation

oai_token-act-pair · gpt-3.5-turbo

references to components or elements of various processes or activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Leilan

-0.65

 ceilings

-0.65

culosis

-0.60

ulner

-0.59

 Predators

-0.58

liam

-0.58

izu

-0.57

ampires

-0.56

 heights

-0.56

 Shap

-0.56

POSITIVE LOGITS

meal

0.86

ridge

0.84

aking

0.74

displayText

0.73

heartedly

0.71

ridges

0.69

dated

0.68

furt

0.68

etsu

0.67

lined

0.65

Activations Density 0.026%

phrases indicating a specific action or event that is part of a broader process or investigation

references to components or elements of various processes or activities

No Comments

No Known Activations