INDEX

Explanations

statistically significant results within text

oai_token-act-pair · gpt-3.5-turbo

phrases indicating statistical significance and relationships in research findings

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

dust

-0.74

roots

-0.74

wire

-0.72

 commons

-0.72

lance

-0.68

åĤ

-0.68

din

-0.67

plates

-0.66

ä¸Ĭ

-0.65

 vows

-0.65

POSITIVE LOGITS

 (âĪĴ

1.13

 detectable

1.04

 statistically

1.00

 heterogeneity

0.96

RESULTS

0.89

 comparing

0.87

 correlated

0.86

 correlation

0.86

 correlate

0.85

 correlations

0.85

Activations Density 0.142%

statistically significant results within text

phrases indicating statistical significance and relationships in research findings

No Comments

No Known Activations