INDEX

Explanations

instances where someone has no issue or difficulty with a given situation or action

oai_token-act-pair · gpt-3.5-turbo

repetitions of the phrase "no problem."

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

ãĥ¼ãĥĨãĤ£

-0.76

ystem

-0.71

byter

-0.70

ModLoader

-0.69

ospel

-0.68

ificial

-0.67

ioxide

-0.66

 Hitch

-0.64

ocaly

-0.63

odium

-0.62

POSITIVE LOGITS

 whatsoever

1.55

 anymore

0.77

nor

0.76

iating

0.69

tips

0.68

 complying

0.68

 reelection

0.67

 anywhere

0.66

 bothered

0.64

 harass

0.63

Activations Density 0.045%

instances where someone has no issue or difficulty with a given situation or action

repetitions of the phrase "no problem."

No Comments

No Known Activations