INDEX

Explanations

specific words related to technology features or specifications

oai_token-act-pair · gpt-3.5-turbo

references to types or categories, particularly in a technical or specification context

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

romeda

-0.90

olulu

-0.72

utical

-0.71

å§«

-0.70

ordial

-0.70

ITNESS

-0.69

yrinth

-0.68

pton

-0.67

nas

-0.67

ernel

-0.67

POSITIVE LOGITS

faces

1.25

face

1.17

ahead

0.85

casting

0.80

etter

0.78

etting

0.75

Script

0.71

 inference

0.68

inker

0.68

oho

0.68

Activations Density 0.018%

specific words related to technology features or specifications

references to types or categories, particularly in a technical or specification context

No Comments

No Known Activations