INDEX

Explanations

references to individuals named "Oro" with varying levels of emphasis

oai_token-act-pair · gpt-3.5-turbo

words related to powerful or influential figures

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

ablishment

-0.77

ivities

-0.74

 Corpus

-0.68

 Extrem

-0.65

ivism

-0.63

Lost

-0.61

Temperature

-0.60

NER

-0.59

nings

-0.58

 Sunshine

-0.58

POSITIVE LOGITS

oro

1.40

zzi

1.07

vernment

1.03

annis

0.89

oros

0.88

eco

0.87

onto

0.85

omon

0.84

heim

0.83

oso

0.81

Activations Density 0.009%

references to individuals named "Oro" with varying levels of emphasis

words related to powerful or influential figures

No Comments

No Known Activations