INDEX

Explanations

terms related to names, specifically those ending with 'ole'

oai_token-act-pair · gpt-3.5-turbo

variations of the word "ole," particularly in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

taboola

-0.65

è¦ļéĨĴ

-0.64

DX

-0.62

Phoenix

-0.61

é¾

-0.61

wcsstore

-0.61

Rap

-0.60

 Rampage

-0.60

CoC

-0.57

HCR

-0.57

POSITIVE LOGITS

ole

1.36

tto

1.11

cules

1.09

tta

1.07

tti

1.05

cule

1.01

lette

0.98

ase

0.89

lla

0.88

cular

0.88

Activations Density 0.007%

terms related to names, specifically those ending with 'ole'

variations of the word "ole," particularly in various contexts

No Comments

No Known Activations