INDEX

Explanations

descriptions of actions or events happening in a sequence

oai_token-act-pair · gpt-3.5-turbo

phrases indicating significant changes or processes

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

Tud

-0.73

Bat

-0.71

Tang

-0.70

Bat

-0.64

 Scots

-0.63

 Philippines

-0.62

Uk

-0.62

Dug

-0.62

Laughs

-0.60

Serial

-0.59

POSITIVE LOGITS

atto

0.92

ItemTracker

0.90

 consisted

0.74

 reportedly

0.74

 eventually

0.72

isphere

0.71

 however

0.71

 moreover

0.70

 housed

0.70

 subsequently

0.70

Activations Density 1.157%

descriptions of actions or events happening in a sequence

phrases indicating significant changes or processes

No Comments

No Known Activations