INDEX

Explanations

Do note that the neuron 4 activates particularly when it encounters the word "quirk" in various contexts

oai_token-act-pair · gpt-3.5-turbo

words related to quirks or oddities, particularly in reference to people or behavior

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

hur

-0.68

rade

-0.64

iam

-0.63

amaz

-0.62

ÙĴ

-0.62

gang

-0.62

sword

-0.60

âĺħâĺħ

-0.59

++++

-0.59

 belt

-0.59

POSITIVE LOGITS

conservancy

0.87

earance

0.82

enhagen

0.81

atcher

0.78

utic

0.75

adem

0.73

opsis

0.69

ensable

0.69

lihood

0.67

 Cipher

0.67

Activations Density 0.050%

Do note that the neuron 4 activates particularly when it encounters the word "quirk" in various contexts

words related to quirks or oddities, particularly in reference to people or behavior

No Comments

No Known Activations