INDEX

Explanations

proper nouns or terms related to software features or updates

oai_token-act-pair · gpt-3.5-turbo

terms related to digital content or file organization

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

aneously

-0.80

lier

-0.77

 exception

-0.75

istically

-0.74

ously

-0.71

activity

-0.68

IOR

-0.67

xious

-0.67

ffer

-0.66

ren

-0.66

POSITIVE LOGITS

peed

1.33

hift

1.28

mith

1.21

pace

1.16

aurus

1.16

pring

1.11

pread

1.10

ystem

1.09

ilver

1.04

ource

1.03

Activations Density 0.209%

proper nouns or terms related to software features or updates

terms related to digital content or file organization

No Comments

No Known Activations