INDEX

Explanations

references to and descriptions of official narratives or reports

oai_token-act-pair · gpt-3.5-turbo

references to authoritative sources or common phrases beginning with "the."

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

apon

-0.97

quit

-0.82

icably

-0.78

rontal

-0.73

ados

-0.72

thood

-0.72

 collide

-0.71

ooth

-0.70

ÃĥÃĤ

-0.69

successfully

-0.69

POSITIVE LOGITS

 latest

1.09

 latter

1.05

 aforementioned

1.04

 same

0.97

 Centers

0.86

 outset

0.86

 extent

0.84

 widest

0.83

 earliest

0.83

 agency

0.82

Activations Density 0.137%

references to and descriptions of official narratives or reports

references to authoritative sources or common phrases beginning with "the."

No Comments

No Known Activations