INDEX

Explanations

words related to locations or names ending in '-ci'

oai_token-act-pair · gpt-3.5-turbo

references to the word "ci."

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

 odds

-0.78

locked

-0.77

stood

-0.77

grounds

-0.76

birds

-0.74

rooms

-0.70

ton

-0.67

*/(

-0.65

footed

-0.65

 BaseType

-0.65

POSITIVE LOGITS

ci

1.08

enza

1.00

Ã¨re

0.97

zona

0.95

ption

0.94

ppo

0.93

ences

0.91

arbon

0.91

pport

0.91

zz

0.90

Activations Density 0.008%

words related to locations or names ending in '-ci'

references to the word "ci."

No Comments

No Known Activations