INDEX

Explanations

keywords related to shortcuts and computing

oai_token-act-pair · gpt-3.5-turbo

references to shortcuts and related terms in the context of computing or navigation

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

IGH

-0.77

endi

-0.74

andum

-0.73

rum

-0.70

omed

-0.69

mos

-0.68

ŃĶ

-0.67

haus

-0.67

osuke

-0.66

league

-0.66

POSITIVE LOGITS

 shortcuts

1.22

 shortcut

1.06

 showc

0.86

éĹĺ

0.74

 pane

0.72

TY

0.72

 hint

0.71

 sled

0.71

 Directions

0.69

 workaround

0.68

Activations Density 0.015%

keywords related to shortcuts and computing

references to shortcuts and related terms in the context of computing or navigation

No Comments

No Known Activations