INDEX

Explanations

proper nouns and technical terms, possibly related to politics, economics, and technology

oai_token-act-pair · gpt-3.5-turbo

specific terms related to measurement and scientific or technical concepts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 pione

-0.96

 eleph

-0.83

Þ

-0.73

 tremend

-0.71

Bun

-0.69

Bak

-0.68

 practition

-0.67

 rall

-0.66

 exting

-0.65

cloneembedreportprint

-0.65

POSITIVE LOGITS

·

0.75

nea

0.72

olesterol

0.71

cible

0.70

º

0.67

ics

0.66

etry

0.64

Ń

0.64

ĺ

0.64

etime

0.64

Activations Density 0.232%

proper nouns and technical terms, possibly related to politics, economics, and technology

specific terms related to measurement and scientific or technical concepts

No Comments

No Known Activations