INDEX

Explanations

words related to locations or place names, particularly those related to infrastructure

oai_token-act-pair · gpt-3.5-turbo

abbreviations and acronyms related to governmental or organizational entities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 kittens

-0.69

oppers

-0.57

 bare

-0.57

 puppies

-0.56

umenthal

-0.55

accompan

-0.55

oppable

-0.55

Reviewer

-0.54

levels

-0.54

itz

-0.53

POSITIVE LOGITS

ividual

1.08

iband

0.81

asonic

0.77

uments

0.77

urities

0.75

pport

0.74

 Micha

0.73

querque

0.72

adow

0.68

entials

0.68

Activations Density 0.062%

words related to locations or place names, particularly those related to infrastructure

abbreviations and acronyms related to governmental or organizational entities

No Comments

No Known Activations