INDEX

Explanations

directives or instructions emphasizing limits or constraints

oai_token-act-pair · gpt-3.5-turbo

phrases indicating limitations or constraints

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

ult

-0.72

chwitz

-0.65

Dru

-0.62

ameron

-0.62

itch

-0.61

ennes

-0.61

BDS

-0.61

ellation

-0.60

umn

-0.60

overty

-0.60

POSITIVE LOGITS

 fraction

0.75

ONE

0.72

ifiable

0.72

 spor

0.70

 finite

0.69

 limited

0.68

 semblance

0.68

 positives

0.68

 provisional

0.67

soever

0.67

Activations Density 0.352%

directives or instructions emphasizing limits or constraints

phrases indicating limitations or constraints

No Comments

No Known Activations