INDEX

Explanations

terms related to negative events and incidents

oai_token-act-pair · gpt-3.5-turbo

terms related to accidents and their investigations

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

raltar

-0.74

agog

-0.65

©¶æ

-0.64

conservancy

-0.63

Camera

-0.62

uably

-0.62

anyl

-0.60

ribes

-0.60

 facult

-0.59

 oppressed

-0.58

POSITIVE LOGITS

 spree

0.95

 happening

0.86

 unfolding

0.85

 unfold

0.80

 occurring

0.78

nings

0.75

 announcement

0.72

 ordeal

0.72

0.69

,.

0.67

Activations Density 0.307%

terms related to negative events and incidents

terms related to accidents and their investigations

No Comments

No Known Activations