INDEX

Explanations

words and phrases related to making judgments or evaluations

oai_token-act-pair · gpt-3.5-turbo

references to the concept of judgment or evaluative decisions

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

href

-0.80

tails

-0.71

tail

-0.67

ieri

-0.66

repeat

-0.63

FORMATION

-0.62

ohyd

-0.61

vae

-0.59

chell

-0.59

auc

-0.59

POSITIVE LOGITS

 judgment

1.29

 judgement

1.22

 judgments

1.07

 Judgment

1.01

naire

0.87

jud

0.87

eering

0.79

 debtor

0.75

 ACTIONS

0.74

al

0.72

Activations Density 0.009%

words and phrases related to making judgments or evaluations

references to the concept of judgment or evaluative decisions

No Comments

No Known Activations