INDEX

Explanations

comparisons of quantitative differences

oai_token-act-pair · gpt-3.5-turbo

comparative phrases indicating levels or conditions relative to the past

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

agy

-0.64

isible

-0.63

ema

-0.62

yi

-0.60

ado

-0.59

ebin

-0.58

 partName

-0.58

SourceFile

-0.57

ocyte

-0.56

ango

-0.55

POSITIVE LOGITS

 usual

1.14

 elsewhere

0.99

 they

0.91

usual

0.88

 ever

0.84

 before

0.81

 originally

0.79

 formerly

0.79

before

0.79

 previously

0.77

Activations Density 0.120%

comparisons of quantitative differences

comparative phrases indicating levels or conditions relative to the past

No Comments

No Known Activations