INDEX

Explanations

phrases related to expressed thoughts or opinions

oai_token-act-pair · gpt-3.5-turbo

instances of the word "that" 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 @ 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

aukee

-0.85

ãĤ´ãĥ³

-0.81

 pione

-0.79

ä

-0.77

iped

-0.77

arest

-0.74

afe

-0.74

ascript

-0.73

amia

-0.71

ãĤĬ

-0.70

POSITIVE LOGITS

's

0.97

 happens

0.95

 wasn

0.91

 translates

0.90

 settles

0.90

 justifies

0.90

 applies

0.89

 doesn

0.89

 proves

0.88

isn

0.88

Activations Density 0.210%

phrases related to expressed thoughts or opinions

instances of the word "that" in various contexts

No Comments

No Known Activations