INDEX

Explanations

proper nouns beginning with "Prim"

oai_token-act-pair · gpt-3.5-turbo

instances of the word "Prim" and its variations in the text

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

INESS

-0.89

lihood

-0.86

IGH

-0.78

UGH

-0.76

isSpecialOrderable

-0.71

IELD

-0.69

DCS

-0.66

UGE

-0.66

TPS

-0.65

 Graveyard

-0.63

POSITIVE LOGITS

itives

1.44

itive

1.23

ordial

1.18

arily

1.00

archs

0.95

acy

0.92

rose

0.92

eness

0.89

 Prim

0.88

atis

0.88

Activations Density 0.015%

proper nouns beginning with "Prim"

instances of the word "Prim" and its variations in the text

No Comments

No Known Activations