INDEX

Explanations

terms related to scientific concepts, especially those involving concentration gradients and medical conditions like concentration camps

oai_token-act-pair · gpt-3.5-turbo

references to "concentration," particularly in scientific or historical contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

mic

-0.76

ĪĴ

-0.75

pher

-0.70

HEAD

-0.69

enic

-0.68

named

-0.64

soever

-0.64

 behalf

-0.62

TERN

-0.62

tell

-0.61

POSITIVE LOGITS

 concentration

1.18

 concentrations

0.93

uations

0.82

anguage

0.81

 concentrated

0.81

emetery

0.80

 Concent

0.78

wagon

0.76

elsius

0.76

uation

0.75

Activations Density 0.011%

terms related to scientific concepts, especially those involving concentration gradients and medical conditions like concentration camps

references to "concentration," particularly in scientific or historical contexts

No Comments

No Known Activations