INDEX

Explanations

strings related to adding or appending information

oai_token-act-pair · gpt-3.5-turbo

references to additions or increments in data or statistics

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 reception

-0.69

 grain

-0.62

 ultr

-0.62

 demo

-0.62

 blade

-0.61

 Siberia

-0.60

 demos

-0.60

 handshake

-0.59

 live

-0.59

fem

-0.59

POSITIVE LOGITS

add

4.15

ADD

2.26

added

1.82

adding

1.74

Add

1.72

addle

1.64

ads

1.58

Add

1.43

ADD

1.38

addy

1.37

Activations Density 0.015%

strings related to adding or appending information

references to additions or increments in data or statistics

No Comments

No Known Activations