INDEX

Explanations

words related to loss or defeat

oai_token-act-pair · gpt-3.5-turbo

words related to losing or negative outcomes

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

è¦ļéĨĴ

-0.59

 Dunham

-0.59

 Express

-0.59

 pronounced

-0.58

 registers

-0.57

 interchange

-0.57

 Alps

-0.55

Lv

-0.55

DOC

-0.55

 Jewel

-0.54

POSITIVE LOGITS

osing

1.07

osed

0.91

pose

0.81

anasia

0.79

oses

0.77

dinand

0.74

allery

0.74

anamo

0.73

oser

0.73

ricanes

0.72

Activations Density 0.006%

words related to loss or defeat

words related to losing or negative outcomes

No Comments

No Known Activations