INDEX

Explanations

explanatory statements or reasons given for various situations or decisions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "explaining" and related concepts of clarification or justification

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

++

-0.65

ļ

-0.63

eco

-0.62

ouble

-0.60

MJ

-0.59

aband

-0.58

cakes

-0.57

orate

-0.57

rose

-0.56

POSITIVE LOGITS

 explaining

3.53

 describing

2.09

 justifying

1.90

 explain

1.86

 explanation

1.77

 explan

1.76

 Explain

1.72

 arguing

1.68

 apologizing

1.68

 illustrating

1.67

Activations Density 0.010%

explanatory statements or reasons given for various situations or decisions

instances of the word "explaining" and related concepts of clarification or justification

No Comments

No Known Activations