INDEX

Explanations

instances of the word "an", particularly when occurring just before a noun or noun phrase

oai_token-act-pair · gpt-3.5-turbo

the articles "an" and "a" 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 @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

ching

-0.71

birds

-0.66

 bras

-0.65

stones

-0.64

verting

-0.63

 dear

-0.62

Topics

-0.61

 Brach

-0.61

 jokes

-0.60

pes

-0.60

POSITIVE LOGITS

alyses

1.02

abolic

0.99

 optional

0.93

 alternative

0.92

 assortment

0.90

 overview

0.88

 independent

0.88

aer

0.88

 orderly

0.87

onym

0.87

Activations Density 0.185%

instances of the word "an", particularly when occurring just before a noun or noun phrase

the articles "an" and "a" in various contexts

No Comments

No Known Activations