INDEX

Explanations

words related to definitions or descriptions of concepts or things

oai_token-act-pair · gpt-3.5-turbo

instances of the word "define" or its variations, indicating a focus on the concept of definition itself

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

athi

-0.86

ience

-0.85

ili

-0.72

ery

-0.69

 Vive

-0.69

ilia

-0.68

Ã£

-0.67

äºĶ

-0.67

inho

-0.67

ublic

-0.66

POSITIVE LOGITS

initions

0.99

 defined

0.80

 defines

0.80

 defining

0.77

 definitions

0.77

 thresholds

0.74

 define

0.73

inition

0.71

 boundaries

0.70

 FORM

0.70

Activations Density 0.027%

words related to definitions or descriptions of concepts or things

instances of the word "define" or its variations, indicating a focus on the concept of definition itself

No Comments

No Known Activations