INDEX

Explanations

commands or requests directed towards an audience

oai_token-act-pair · gpt-3.5-turbo

phrases indicating announcements, actions, or statements related to companies or projects

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

osi

-0.74

erence

-0.73

anging

-0.67

 reciproc

-0.67

ã

-0.65

efe

-0.65

ischer

-0.64

ittal

-0.64

otta

-0.63

oppy

-0.63

POSITIVE LOGITS

Ltd

0.94

Inc

0.86

 Gleaming

0.84

LLC

0.79

iT

0.77

Plugin

0.77

 Limited

0.76

Premium

0.75

LTD

0.74

Offline

0.74

Activations Density 0.801%

commands or requests directed towards an audience

phrases indicating announcements, actions, or statements related to companies or projects

No Comments

No Known Activations