INDEX

Explanations

statements where someone is explaining or elaborating on a subject

oai_token-act-pair · gpt-3.5-turbo

phrases that start with "to" followed by a verb indicating an action or explanation

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

soDeliveryDate

-0.86

 Compass

-0.66

 Cube

-0.66

Dash

-0.64

 Ended

-0.64

 Crate

-0.63

orno

-0.61

inventory

-0.61

avorite

-0.61

Enlarge

-0.61

POSITIVE LOGITS

 dismantle

0.88

 achieve

0.84

 liberate

0.80

 explore

0.79

 establish

0.76

pless

0.76

 erase

0.76

Reloaded

0.76

 fulfill

0.76

 elaborate

0.75

Activations Density 0.070%

statements where someone is explaining or elaborating on a subject

phrases that start with "to" followed by a verb indicating an action or explanation

No Comments

No Known Activations