INDEX

Explanations

statements or conclusions presented by reports

oai_token-act-pair · gpt-3.5-turbo

statements expressing findings or conclusions from reports

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Himself

-0.79

alogue

-0.73

ãĥ¡

-0.73

ãĥ¬

-0.71

achable

-0.68

())

-0.68

à¸

-0.67

ãĥ¯ãĥ³

-0.67

ãĥİ

-0.67

OK

-0.67

POSITIVE LOGITS

 scathing

0.82

 irregularities

0.74

 findings

0.74

 gaps

0.70

enhagen

0.69

 anomalies

0.69

 anecdotal

0.68

 systematic

0.68

 economists

0.67

 disparities

0.67

Activations Density 0.288%

statements or conclusions presented by reports

statements expressing findings or conclusions from reports

No Comments

No Known Activations