INDEX

Explanations

phrases indicating accumulation or final result

oai_token-act-pair · gpt-3.5-turbo

phrases related to accumulation and sums

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

nor

-0.68

ellen

-0.64

CDC

-0.63

rongh

-0.59

inski

-0.58

ording

-0.57

erved

-0.55

eah

-0.55

 Quit

-0.54

arta

-0.54

POSITIVE LOGITS

 nicely

0.98

 into

0.94

 beautifully

0.84

to

0.83

 phenomen

0.77

 horribly

0.76

 neatly

0.75

ect

0.75

 disastrous

0.75

 seamlessly

0.72

Activations Density 0.169%

phrases indicating accumulation or final result

phrases related to accumulation and sums

No Comments

No Known Activations