INDEX

Explanations

abbreviations and proper nouns

oai_token-act-pair · gpt-3.5-turbo

terms related to emergency response or regulatory entities

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

ĸļ

-0.82

ãĥĥãĥī

-0.81

luaj

-0.76

raints

-0.72

owler

-0.71

 Strauss

-0.70

 Mayweather

-0.65

emonium

-0.64

 Singer

-0.64

hedral

-0.64

POSITIVE LOGITS

BACK

1.13

ICA

1.11

IES

1.08

RY

1.07

WORK

1.06

ATOR

1.06

TAIN

1.05

FIELD

1.05

ANGE

1.04

LY

1.04

Activations Density 0.013%

abbreviations and proper nouns

terms related to emergency response or regulatory entities

No Comments

No Known Activations