INDEX

Explanations

phrases related to updating and informing

oai_token-act-pair · gpt-3.5-turbo

phrases related to providing updates or new information

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

WOOD

-0.94

inki

-0.78

SPONSORED

-0.76

PLIED

-0.74

baum

-0.72

NetMessage

-0.69

Young

-0.68

ottest

-0.67

zees

-0.67

ptin

-0.66

POSITIVE LOGITS

 rosters

0.69

 revised

0.68

undown

0.68

 menu

0.68

facade

0.67

 layout

0.67

 existing

0.66

 version

0.65

 dere

0.64

 timeline

0.63

Activations Density 0.333%

phrases related to updating and informing

phrases related to providing updates or new information

No Comments

No Known Activations