INDEX

Explanations

emotive or evaluative terms indicating high value or satisfaction

oai_token-act-pair · gpt-3.5-turbo

important and significant nouns or concepts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

İĭ

-0.86

ancies

-0.74

tnc

-0.69

oneliness

-0.67

rocal

-0.66

zzy

-0.65

conservancy

-0.64

ounded

-0.63

tti

-0.62

 unfocusedRange

-0.60

POSITIVE LOGITS

 imaginable

1.26

 ever

1.10

 EVER

1.00

 conceivable

0.87

 available

0.78

ever

0.75

 possible

0.74

 confronting

0.73

ally

0.72

Ever

0.71

Activations Density 0.224%

emotive or evaluative terms indicating high value or satisfaction

important and significant nouns or concepts

No Comments

No Known Activations