INDEX

Explanations

references to investigations or potential collusions involving Russia, particularly related to the 2016 US presidential election

oai_token-act-pair · gpt-3.5-turbo

mentions of Russia and its involvement in US politics

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

 Dart

-0.77

dash

-0.76

Thom

-0.74

prototype

-0.74

Label

-0.72

}}}

-0.71

jj

-0.70

jc

-0.69

ulton

-0.69

draw

-0.68

POSITIVE LOGITS

 dossier

0.91

 collusion

0.89

 blackmail

0.89

 meddling

0.88

 infiltrated

0.86

 interfered

0.85

 disinformation

0.84

 intelligence

0.83

 invaded

0.83

afort

0.82

Activations Density 0.034%

references to investigations or potential collusions involving Russia, particularly related to the 2016 US presidential election

mentions of Russia and its involvement in US politics

No Comments

No Known Activations