INDEX

Explanations

phrases related to specific individuals, particularly regarding their achievements or mentions of their names

oai_token-act-pair · gpt-3.5-turbo

the name "Reagan" in various contexts and forms

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

upon

-0.78

İĭ

-0.71

hovah

-0.64

 Pyth

-0.64

bred

-0.63

Peg

-0.63

inished

-0.61

ĵĺ

-0.60

grades

-0.59

 bour

-0.58

POSITIVE LOGITS

igans

1.03

furt

1.03

zeb

0.91

arde

0.88

za

0.88

omics

0.86

ovy

0.85

azi

0.85

osity

0.84

agan

0.84

Activations Density 0.012%

phrases related to specific individuals, particularly regarding their achievements or mentions of their names

the name "Reagan" in various contexts and forms

No Comments

No Known Activations