INDEX

Explanations

technical terms related to conducting tests or experiments

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "test" and its variations, indicating a focus on testing processes and results

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 Gothic

-0.75

leans

-0.66

usterity

-0.66

Fol

-0.65

Cro

-0.65

spoken

-0.62

 Soup

-0.62

ilingual

-0.62

rising

-0.62

 Hust

-0.60

POSITIVE LOGITS

 tests

0.89

 testing

0.86

imony

0.82

wcsstore

0.81

 labs

0.81

 testers

0.79

 prototypes

0.78

oys

0.77

rador

0.77

 sessions

0.76

Activations Density 0.110%

technical terms related to conducting tests or experiments

occurrences of the word "test" and its variations, indicating a focus on testing processes and results

No Comments

No Known Activations