INDEX

Explanations

instances where the term "new" is used

oai_token-act-pair · gpt-3.5-turbo

instances of the word "new" 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 @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

iffs

-0.64

 bond

-0.62

 decomp

-0.60

folios

-0.60

 ****

-0.58

 Skydragon

-0.58

 Sunder

-0.58

 unspecified

-0.57

 Junk

-0.56

ository

-0.56

POSITIVE LOGITS

nor

0.95

 anymore

0.89

 Canaver

0.75

UFC

0.74

quist

0.74

theless

0.71

grass

0.69

til

0.64

ãĤ´

0.64

ãĥ¼ãĤ¯

0.64

Activations Density 0.055%

instances where the term "new" is used

instances of the word "new" in various contexts

No Comments

No Known Activations