INDEX

Explanations

words related to updates, changes, and enhancements in a technical context

oai_token-act-pair · gpt-3.5-turbo

instances of the word "now" indicating updates or changes

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

quickShipAvailable

-0.96

str

-0.72

ogi

-0.70

 repetition

-0.69

achev

-0.67

sym

-0.67

utherland

-0.67

wig

-0.66

ELD

-0.66

tnc

-0.66

POSITIVE LOGITS

adays

1.28

 reversible

0.88

here

0.86

 defunct

0.80

 permanently

0.74

 correctly

0.73

 officially

0.73

 obsolete

0.70

 happily

0.69

 Offline

0.67

Activations Density 0.062%

words related to updates, changes, and enhancements in a technical context

instances of the word "now" indicating updates or changes

No Comments

No Known Activations