INDEX

Explanations

names associated with political figures, particularly related to the Trump campaign

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs24576-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.8.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

ĸļ

-0.82

ï¸

-0.81

 Curiosity

-0.75

Parser

-0.72

Pu

-0.71

 Pixie

-0.71

Redd

-0.69

Pax

-0.68

 Fiji

-0.68

 à¨

-0.68

POSITIVE LOGITS

afort

1.33

 indicted

0.90

 Manafort

0.88

aundering

0.74

aunder

0.73

agements

0.73

 seizures

0.72

uchin

0.71

opoulos

0.71

 misled

0.71

Activations Density 0.019%

names associated with political figures, particularly related to the Trump campaign

No Comments

No Known Activations