INDEX

Explanations

terms related to 'nerd' or 'nerdy'

oai_token-act-pair · gpt-3.5-turbo

references to 'nerd' and its variations, indicating a focus on the nerd culture or identity

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

achusetts

-0.69

Mary

-0.68

Multiple

-0.64

 FISA

-0.64

 newsletters

-0.61

CPC

-0.60

DVD

-0.59

 Corruption

-0.59

 Cipher

-0.58

 Antar

-0.57

POSITIVE LOGITS

linger

0.95

pei

0.89

geon

0.89

ding

0.86

dy

0.85

uling

0.83

azer

0.82

wart

0.82

vana

0.81

icer

0.81

Activations Density 0.017%

terms related to 'nerd' or 'nerdy'

references to 'nerd' and its variations, indicating a focus on the nerd culture or identity

No Comments

No Known Activations