INDEX

Explanations

pieces of computer code or programming syntax

oai_token-act-pair · gpt-3.5-turbo

numerical data or structured code elements

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

pmwiki

-0.94

ometimes

-0.84

ospace

-0.82

 infographic

-0.79

thood

-0.78

 lett

-0.74

indal

-0.74

 shorth

-0.72

ãĥ¼ãĥĨ

-0.72

ulhu

-0.71

POSITIVE LOGITS

 kernel

0.79

 DEBUG

0.78

 Device

0.76

\\

0.75

 ERROR

0.74

API

0.74

 DHCP

0.73

 Kernel

0.72

:[

0.72

 Driver

0.71

Activations Density 0.162%

pieces of computer code or programming syntax

numerical data or structured code elements

No Comments

No Known Activations