INDEX

Explanations

terminology related to frameworks or structures

oai_token-act-pair · gpt-3.5-turbo

references to structural systems or models

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

hib

-0.80

dream

-0.75

cial

-0.68

prom

-0.66

igi

-0.66

hiro

-0.66

rooms

-0.65

aqu

-0.65

 grandson

-0.63

hi

-0.63

POSITIVE LOGITS

 framework

1.08

framework

0.87

 frameworks

0.86

 Framework

0.80

 backbone

0.71

 guidelines

0.70

 coales

0.68

ahime

0.68

yip

0.67

 principles

0.67

Activations Density 0.009%

terminology related to frameworks or structures

references to structural systems or models

No Comments

No Known Activations