INDEX

Explanations

references to the concept of 'tipping'

oai_token-act-pair · gpt-3.5-turbo

terms related to tipping points or thresholds in various contexts

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

Interstitial

-0.75

NCT

-0.73

 Purpose

-0.72

venants

-0.71

inelli

-0.68

 Clinic

-0.67

 Clubs

-0.66

ively

-0.66

 Manual

-0.65

ee

-0.65

POSITIVE LOGITS

 tipping

0.99

 iceberg

0.98

 tipped

0.81

ratulations

0.74

iod

0.73

archy

0.71

olia

0.70

uple

0.70

achy

0.70

 earners

0.70

Activations Density 0.015%

references to the concept of 'tipping'

terms related to tipping points or thresholds in various contexts

No Comments

No Known Activations