INDEX

Explanations

the word "aur"

oai_token-act-pair · gpt-3.5-turbo

occurrences of the name "Saur" and its variations

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

 Shroud

-0.69

 Sparkle

-0.68

 Winchester

-0.65

FU

-0.65

 Perspective

-0.63

 goto

-0.63

 Grid

-0.62

LINE

-0.62

 Proced

-0.60

 Reloaded

-0.60

POSITIVE LOGITS

aur

1.08

issan

1.07

anamo

0.91

idon

0.90

rible

0.88

ricane

0.88

uses

0.87

orst

0.85

isance

0.83

itor

0.83

Activations Density 0.009%

the word "aur"

occurrences of the name "Saur" and its variations

No Comments

No Known Activations