INDEX

Explanations

edits or comments within a text

oai_token-act-pair · gpt-3.5-turbo

instances of edits or modifications made to a document

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

natureconservancy

-0.71

 Antar

-0.67

@#&

-0.62

 preservation

-0.61

llah

-0.60

gging

-0.60

 ultras

-0.59

fund

-0.58

 beads

-0.58

 Leone

-0.57

POSITIVE LOGITS

orship

1.05

ional

1.03

ors

0.97

OR

0.95

ories

0.90

orable

0.88

arro

0.88

ored

0.87

ORS

0.86

rator

0.86

Activations Density 0.024%

edits or comments within a text

instances of edits or modifications made to a document

No Comments

No Known Activations