INDEX

Explanations

phrases related to negative events or occurrences

oai_token-act-pair · gpt-3.5-turbo

terms associated with death and significant life events

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

kW

-0.73

VERTIS

-0.69

 linear

-0.66

 Kers

-0.65

cest

-0.64

 Greenberg

-0.62

GUN

-0.61

olic

-0.60

 Berman

-0.59

oros

-0.58

POSITIVE LOGITS

 demise

0.84

ishment

0.78

 announcement

0.74

 proclamation

0.73

 woes

0.73

stretched

0.72

 spree

0.71

 doomed

0.71

 mate

0.70

 journey

0.70

Activations Density 0.251%

phrases related to negative events or occurrences

terms associated with death and significant life events

No Comments

No Known Activations