INDEX

Explanations

numbers or version identifiers

oai_token-act-pair · gpt-3.5-turbo

numerical values and punctuation in a structured format

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

owship

-0.74

 boarding

-0.66

cheon

-0.63

WOOD

-0.62

 wedd

-0.62

 hawk

-0.61

 boun

-0.61

iage

-0.60

lean

-0.59

Arthur

-0.59

POSITIVE LOGITS

0.77

 Centauri

0.76

[+

0.73

 Introduction

0.72

ually

0.72

gran

0.71

0.70

umbing

0.69

0.67

rss

0.67

Activations Density 0.159%

numbers or version identifiers

numerical values and punctuation in a structured format

No Comments

No Known Activations