INDEX

Explanations

verbs related to actions or changes that occur as a result of a previous event

oai_token-act-pair · gpt-3.5-turbo

phrases related to consequences and their effects

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

rss

-0.69

icip

-0.69

 kidding

-0.67

arently

-0.66

aspers

-0.66

 sparing

-0.64

maxwell

-0.64

isco

-0.63

theless

-0.62

itches

-0.62

POSITIVE LOGITS

 Legions

0.71

 Mankind

0.70

GW

0.66

SpaceEngineers

0.65

fall

0.63

 Yanuk

0.62

 Gamergate

0.62

 Judgment

0.61

 besie

0.61

 Magicka

0.61

Activations Density 0.857%

verbs related to actions or changes that occur as a result of a previous event

phrases related to consequences and their effects

No Comments

No Known Activations